- The exit requires as little as 10% selling their coins in the previous majority ledger, which effects a huge decrease in the value of the remaining coins there due to the negative wealth effect, while increasing the value of their new ledger by approximately the equivalent amount of positive wealth effect. Thus the small minority of the "rogues" can enrich themselves in the expense of the majority by so doing, creating a new majority ledger in terms of marketcap.
Very much interested in hearing rebuttals which take into account the human psychology and stay on the ground of voluntarism

If your main point is how few "rogues" (~10%) it takes to switch to a new majority ledger, I think you may be neglecting the arbitrage opportunities that would create. Arbitrage seems to undo the cascade effect of the small exchange float, with arbitrageurs profiting from that market inefficiency.
That is, if the scenario is such that 10% of investors are motivated to switch their portfolio allocations (whether slowly or quickly), then by hypothesis the remaining 90% are
not motivated to change their portfolio allocations. But the actions of the 10% result in the portfolios of the 90% changing anyway, against their will.
For example, suppose the 90% only wanted to hold a "tiny bit" of their crypto-wealth as LTC and all the rest as BTC, but now the ongoing price change has left them with say "5x a tiny bit" of LTC and a little less BTC.
Insofar as the 90% are mostly strong hands,* not price chasers, they will react by bringing their portfolios back in line. That means they will do the opposite to what the 10% are doing, but with 9x the force: sell their allocations in the minority ledger, now for a giant premium, and buy more BTC at these cheap prices. I believe this negates the "small float" issues of having only a tiny amount of each coin available on exchange at any one time (not to mention that when prices move drastically a lot of coins [and fiat] come out of hiding).
In the end it seems like the price will balance out as one would expect, with the market cap of the new ledger being about 10% that of the Bitcoin ledger, as I think Litecoin once was. The price will of course overshoot to the downside temporarily depending on how fast the switch happens and how weak the average hands are (but exhausting the coins on exchange and thereby bringing coins out of hiding means awakening the Smaug-level strong hands). And any overshoot is yet a further opportunity for arbitrage to further entrench the majority ledger investors who have the strongest hands.
So I tend to think "it all boils down to normalcy" with this one. That is, 10% switching just means 10% switching. If they switched to Litecoin, for example, that would put its market cap at something like 10-15% of Bitcoin's, and for lesser-valued coins a bit closer to 10%.
*
Which is one of the nice things about having a mature ledger that has used years of extreme volatility to throw all but the strongest hands off the bronco