Unless you've chosen a poor sample of the evidence you're familiar with, your opinion is not going to stop anyone from following their fatuous curiosity, here. The historical cases you refer to seem a couple orders of magnitude more fraught with the spooks of subjective indignation than anything anyone in this community would propose. When an analytic philosopher looks at these things they don't see decision procedures that should have worked in theory but failed, they don't see decision procedures at all, they see disagreements in waiting.
I agree that any morally loaded criterion for deciding land reallocations is going to trip over the subjectivity of morality as we know it, especially in a system that's explicitly designed to support the sovereignty of diverse groups. I believe we can at least come up with a negotiation procedure that returns immediate, unambiguous results that do a pretty okay job of cleaning up vacated territories.
I'll call this one Simultaneous Haggle Reallocation.
Let's say that in each term, each state must submit a preference ordering on the areas just outside their border, in neighboring states, and an ordering on the areas just inside their border. The outside list describes the places they'll take if their population increases in proportion to their neighbors, the inside list is the places they'll lose if their population decreases, all in order of their desire to hold them. The top elements of the inside list will be the areas the state most wants to keep. The top of the outside list will be the areas they most want to take. If there is a mutually agreeable way forward to be made, an area they're happy to lose that their neighbor very much wants, or an area they wont part with for cultural reasons that their neighbor doesn't share, that is the trade that will be made.
Kind of unfortunate though.. Above, I provided a formula that assumes an objective(or at least shared) measure of what constitutes habitable land, or, in a more sophisticated implementation; a measure of the value of the land per acre. The more the archipelago agrees on the relative value of land, the more often the states' preference orderings will mirror each other. Much of the time, then, Simplistic Simultanious Haggling as I've defined it would just revert "reallocate at the borders at random(possibly with smoothing) since there's clearly no mutually agreeable way to settle this".
It would be fun to run some simulations of this and see what kind of games emerge.
Historically, the evolution of government systems was mainly driven by violence, with invasions and revolutions being the principal agents of selection process. The rules of the game were predetermined by our environment - land was a limited resource, for which our ancestors had to compete, if only to ensure the survival of their descendants.
The 20th century introduced a game changer. As agricultural productivity in developed countries rose by orders of magnitude and natural population growth practically came to a halt, possessing a large territory stopped being a necessity. Countries with little arable land, ultra-high population density and no natural resources can now not only feed their population, but also achieve top living standards. These changes may open a fundamentally different route for societal evolution – one that would not be based on violence or compulsion.
A small thought experiment - imagine what would happen if central governments cede most powers to smaller territorial units:
Unfortunately, there are serious obstacles to the successful implementation of this idea:
Do you think these problems are solvable?