The main reason I don’t think this solution robust is because it relies on just a number of individuals to monitor LTV ratios, it is done on a weekly basis, and requires those individuals to decide what they consider the market price for COW to be (oracle).
Since this is all public and in light of the relatively thin liquidity for COW, this could be easily gamed. Theoretically by a solver that would like to liquidate their competition.
Having said that, and as I already mentioned above, I think CoW DAO should be open to small scale experiments to try and drive new ideas forward (as long as the risk is managed)