I think a lot of people who speak of MMI don't believe in the possibility of the kind of objective judicious ruling necesary for Free Kriegsspiel.
My view - based on a combination of experience with RPGing and knowledge of Prussian military culture and upper-class culture more generally - is that it depends on combining (i) a fairly narrow basis of fiction/shared imaginary space from which rulings are going to be derived, with (ii) a high degree of shared understanding of the nature and implications of that fiction/SIS.
I think you can see both things at work in the formative period of D&D. The shared fiction is predominantly
rather Spartan dungeons or
ecologically and topographically rather abstract wildernesses. This is the narrow basis I mentioned. And you can see very strong emergent and iterative cultural understandings of what is or isn't possible, a fair "move", etc within that space - Gygax's DMG is notoriously replete with these, which is what makes bits of it so hard to make sense of to readers who weren't part of that shared culture. Examples include: his discussion of how to manage the passage of time, which assumes without stating that the campaign world is being run for multiple groups multiple evenings per week; his discussion around what is appropriate for non-Monty Haul treasure placement, which has to be reconciled with his XP tables and the idea that 10 magic items should be a genuine
limit for a paladin; the obsession with concealed pits as traps and the rules for detecting them, but the relative absence of assassins and the relative unclarity in how the surprise rules should work when one side is setting up an ambush; etc.
My own view is that once the campaign world - talking now not just about background colour and "Gygaxian naturalism" but about the actual subject matter of play - becomes anything like as rich as the real world (and Traveller and Runequest are the earliest RPGs I know of to try and present such gameworlds) then the feasibility of free kriegsspiel adjudication rapidly diminishes.
When mainstream D&D play entered this sort of period is hard to establish with any confidence, for me at least. Tracy Hickman's Desert of Desolation modules are often held up as being early examples of "story"-driven modules, but when I was able to pick them up second-hand a few years ago and have a read of them, they struck me as very dungeon-crawly with a bit of a puzzle-solving overlay. So in my thinking it still comes back to Dragonlance - if that is going to be played not as a dungeon crawl to beat a black dragon but as a genuine "story"-driven experience then I think the Free Kriegsspiel possibilities drop away. No matter how much backstory there is about the Tanis-Kitiar relationship, I don't think there can be objective Free Kriegsspiel determinations of whether or not she would be willing to kill him on the field of battle. There's no "objective" understanding of human emotions and emotional responses that will allow the GM to decide that, and that will bring it within the field of "knowable" prospects for the players. The difference from a covered pit trap, in these respects, could hardly be greater!
They seem to live in a bit of a nightmare world where GM rulings and even reality itself are forms of arbitrary and incomprehensible tyranny.
I don't know about
tyranny; but if the reality of human emotions was as non-arbitrary as I think it needs to be for Free Kriegsspiel adjudication to work, then we would have far fewer songs, poems and rancorous relationship breakups!
If one thinks about literal Free Kriegsspiel, the main emotional factor is morale. But that is not handled by attempting to determine the emotional reactions of any single figure: it's handled by imposing "population"-level generalisations grounded in a shared experience of how those populations respond. Clearly even some of those experiences can produce false population-level conclusions: it seems likely that French adjudicators of Free Kriegsspield would have rated morale as too high a factor in relation to infantry success in contexts of "machine"-warfare; and likewise that many pre-WWII adjudicators would have rated civilian morale against terror bombing as far more likely to break than history has revealed to be the case.
But once we get to single figures, and how they would respond to former loves, whether they have to go to a meeting at a teahouse or just want to take some downtime there,
how they might respond to an SOS signal, etc - well, I'm very sceptical that Free Kriegsspiel methods of adjudication are applicable.