My personal tinfoil..... some of it is textually based, and I shall try to distinguish between what I believe is (suggested by) canon and what is my own wild speculation. Here goes....
Regarding a "self-sufficient side". Yes, it does seem that by default a side has to go to war to maintain its units. But why is this? I think that cities are always popping units, and even if a unit takes longer to pop it will still cost more upkeep than a unit that you can pop multiple groups of in a shorter amount of time. So, regardless of whether you spend the next 30 turns popping a "royal" or "common" warlord or 3-5 sets of stabbers/slashers, you gradually need more and more rations/money to keep your troops from disbanding. If you are always disbanding troops, maybe your sides loyalty gradually decreases. After all, who knows who will be next in this great downsizing? So, either you farm, harvest, manage your cities, use a moneymancer, conquer and raze other cities..... it has been stated that a large portion of a sides strategy is based upon what casters they have. I personally think this is because a large part of Erfworld "strategy" is a overlord thinking about what units they have, and what their cities will pop, how it will affect their upkeep, and what they need to do to have their units earn their upkeep.
Charlie had a sweet deal going, being an isolated "bubble kingdom", one or many things which has been considered "impossible" as far as everyone in Erfworld knows.... But certain types of units also seem to be born with different types of knowledges. For example, it takes a special type of units, I believe a warlord/overlord or a unit capable of leading a stack (like certain archons and all casters) to even be able to see another units stats. Also, Wrigley told us that unled units felt a certain kinship, so their loyalty may well be affected by wanton disbanding of those who have sat there and led pointless lives, draining their sides treasury for turn after turn. And Vurp has told us that not all natural allies trust their commanders. They, or at least Vurp, think it is usually best to remain loyal because that is their best chance for strength. Barbarian Jill had a barely larger than 1700 shmucker purse, so how much can a feral hobgobwin chieftan support? And how likely would he be to find a new tribe to support them? Would they be more likely to die as a neutral random encounter to some established sides scouting groups if they broke alliance and had nowhere to go? These are all useless speculations on my part, but maybe useful ones to consider.
So, basically no one has figured out how to survive without being at the mercy of the game mechanics yet and everyone is scared shitless of disbanding, pardon my french. I could go into more detail if necessary, but basically units without leadership fear being "downsized" or "harvested", having their city razed for upkeep, etc. if their king sucks. Nobles, casters, warlords, heirs etc. can double-deal, and a king must constantly fight a seemingly endless battle against ever increasing upkeep costs and city management.
My basic premise behind this city management is that the more cities you get, the fewer shmuckers each one produces, what Parson meant by the "diminishing shmuckers" point. Each additional city keeps producing troops at the same rate, which means that maybe you could overwhelm smaller sides my sheer production, but you get less and less shmuckers and maybe forage out the areas you can reach/effectively scout and have one Chief Warlord command effectively. This turns into a problem when you are facing 10-15 smaller sides at once (one reason Parsons gamer friends could challenge him... say Parson loses all four Arkentools in a colossal battle and each of his friends spins to a new side with one of the tools? Then they have the resources of four sides and four tools versus Parson, limited by the fairly simple, hard to exploit, turn based mechanics of Erfworld, where he has already showed the denizens many of his fancier moves? Seems like a scary situation for Parson to be in....)
Regarding high magic and caster links, we'll get back to self-sufficient sides in a second (I think there are two, Charlie and the MK)..... Erfworld doesn't seem like a very "scientifically minded" place to me. What I mean by this is that, outside of the magic kingdom, it doesn't seem like that many people take time to debate theory.... And even inside the MK Sizemore is considered unusual for wanting to know about many different disciplines, even though he is terrible at casting outside of his discipline (without a scroll). Even royals might mostly just study famous battles.... who says someone who isn't a Mathamancer of some sort can even have a sound grasp of probability? We know that there are at least three hidden stats, but I believe that klog from the first book indicated there are many more. Maybe one that lets Jillain survive so many mount-dives and dangerous deeds? Maybe one that influences a units perception of fate, or luck (natural predictamancy, natural luckamancy, stealth as natural foolamancy for scouting specials, natural mathamancy for smarter warlords, or hippiemancer for diplomatic ones?). Parson mentioned that a lot of the history books had things like "high scores" instead of deeds, or specials, or tactics. Maybe different regions produce units with different specials, or different sides do, everywhere has its own concentrations of natural "whatevermancy".
If I am correct in assuming that hidden stats are more common in Erfworld that, say "Hits, Attack, Defense, Move" and a special or two, and a lot of it is influenced by "magic", which Parson describes as sounding like a lot of the fake crap that people believed before anyone really understood that you have to poke into things vigorously and thoroughly to arrive at a any semblance of accuracy, it makes sense that a LOT of Erfworlders are very, very ignorant. They know their own specialty, and nothing else, and think of almost nothing else. Parson is not "THE PERFECT WARLORD", merely a tactical genius who can synergize a lot of different strategic elements to come up with a victory. Wanda referred to the spell as one that would summon a Perfect Warlord, both to Stanly (who she needed for the money )and to Sizemore (who isn't in on things). But, when she first mentioned it to Stanly all she said they needed was a tactical genius. To reiterate.... YES, Wanda did tell Stanly it would summon a PERFECT Warlord. But all WANDA really WANTED was A TACTICAL GENIUS WHO COULD THINK HER OUT OF THE HOPELESSLY SCREWED UP SITUATION STANLY HAD THEIR SIDE IN. (All she was looking for was a MIND).
So, I think two sides have some semblance of self-sufficiency. Charlie, who runs his Telecommunication business. He gets more money by a) spreading his archons out so he can spy on people in order to b) intervene when they are at their lowest point so he can c)make it seem like it is a good idea for them to pay him money so he can d) expand his fleet or archons to spy on even more sides which will allow him to e) play everyone against each other (creating problems for our clients creates business) in order to make them smaller which might f) create more and more small sides for Charlescomm to deal with as he encourages countries to kill each other off secretly which would eventually mean that g) Charlie could, at some point, have enough Acrhons spread out throughout the world and veiled to simultaneously or over the course of a few turns take over every last side in Erfworld, after he had played them against each other and whittled them down. Arkentools seem to be freakin awesome.
As for the MK, we have no idea what Tri-mancer links can do.... and just because one Erfworlder in one place used a Tri-mancer link it does not mean that any other Erfworlder will ever figure it out. Recall that Erfworld if not a scientifically inclined place. Why would opposing sides tell each other about what they could do with a tri-mancer link, in the rare occasion that a side has a thinkamancer, and two other casters that they want to put into a link, and a ruler who has given much thought to casting and what he could do with a tri-mancer link, or a warlord who has done the same. Extremely tinfoil hat theory ahead here, but this is as good as any other explanation for The MK, in my view..... The MK is an island, right? So, lets say that at the beginning of time this was a capital city that didn't have anywhere to war against or anywhere to expand to, and it is slowly becoming more and more crowded and having to disband its units. So, it comes up with the solution to try magic to fix its problems! One way or another it has to experiment around, disband some units, work its farms, or a dittomancer, or whatever, but the end result is they create a tri-mancer link that links this kingdom to every other kingdom that will be created, that accept alliance with the Magic Kingdom, but the only units that can enter the portal are casters. Maybe natural casters are included (and if you haven't read the "Parson as a natural Signamancer" section of the Erfwiki I highly recommend it (it is a type of hippimancer concerned with [maaaybe] seeing things as they really are)) but it just isn't tried because who knows if you have a natural 'mancy and you might disband!.
So, everyone allies with TMK because every caster can go there, for free, all the time, and trade for valuable scrolls, links, and services. There are debates..... but the primary purpose for the existence of TMK is commerce. Those casters are expensive! Again recall that even for Jillain's purse of almost 1700 it would have been ludicrous to afford a caster (and interestingly Parson's upkeep is over a thousand.... at level two). TMK "rents out" Casters, almost link Charlie and his archons, except that a caster can choose to permanently join a side, if that sides overlord is willing to pay their upkeep. Moneymancers take payments from whoever bought TMK's services, then they do their mojo and send it to the treasury, where it will pay for the upkeep of all the casters. Meanwhile, they convert it so a static currency based on how many Shmuckers were added to TMK's treasury. This allows all of the casters to trade with each other, and gives them incentive to do something with all that "juice" they get every day. Which leads me to believe they'll have tried many different ways to "store" or "channel" it, if the moneymancers want them all to pay their upkeep efficiently and business is slow for a few turns.
FInally..... I think we are going to see more "nukes" in this story, and you shouldn't be angry at them, like the spell that shall not be mentioned. Here's the thing.... we have no idea how most magic works, or hiw tri-mancer links work. We've been told 4 mancer links don't work, but that doesn't mean 5 mancer links don't work..... x-mancer - thinkamancer - x-mancer - thinkamancer - xmancer, for example. Furthermore, while we don't necessarily know that Charlie has an unmatched command of Thinkamancy (only that he has one Maggie cannot hope to match....as an Adept), I think it is fair to assume he does, due to the friggin artifact. Seeing as he frequently highers casters for his own purposes, and seemingly intuitively guessed how Parson pulled off the volcano trick, I think Charlie has tried to play around with every form of casting in the book. I think he has tri-mancer spells prepared for contingencies out the wazoo. I think he knows most of what there is to know about linking up some casters and using them to pull off some crazy shit. How much will we see of this? That has yet to be seen.
Also, ohmygodwhatamonster of a first post. Late at night, while slightly drunk.
Also also, Oberon, while you occasionally have an interesting point, I think you are largely a dick, and you fail logic forever. Feel free to not discuss.
(Sorry, but caps because I felt that portion might be nitpicked. Still probably will be)