And onwords to discussing the desired functionality!
That’s a great analysis of pros and lingering issues with coin-holding-time weighting (as opposed to the current implementation, which is weighted only by coin holding at a single moment in time).
This is related to @vbuterin’s analysis On Collusion, and Placeholder’s analysis of How Much Does A Crypto-Vote Cost? (h/t @mlphresearch @cburniske). Both make the point that given efficient lending markets, the weight of the vote is proportional to the amount of coins times the “lockup” time.
Unfortunately, by their analysis, the concept of “stake” and “skin in the game” (which is essential to your rationale) is untenable. Given efficient lending markets, the cost of a vote is exactly the interest the voter paid to borrow, or (equivalently), the opportunity cost that the lender paid by not lending. There is no inherent “stake” in the form incentive-aligning exposure to coin price changes.
Separately, there are questions on how to define the holding/lockup period for coin-holding-time weighting. Especially if efficient lending markets are not accessible to all users, there are major limitations/costs in the functionality of the “locked” funds.
There are also circumvention issues raised by @Matthewdgreen.
So… What’s the desired weighting rule, given all of these?
Last but not least, there’s the question of what should be reported vs. kept private. Because if we take a weighting rule seriously, and optimize privacy subject to that, then we should only expose the final tally. But if we try to follow @zooko’s “coin holders, large and small” rallying cry, then… I have no idea what that even means but it would probably imply revealing more than a final weighted tally.