They were using it for grant matching; the grant pool was raised off-chain from donors interested in the betterment of the ecosystem, and then the QF rounds are used to determine how to distribute those grant funds. In that sense, it’s a little different than using it for an up-or-down vote on a specific proposal, but I think that irrespective of the application, without sybil resistance you have a situation where someone willing to split up their funds to make many small voting positions can readily influence the outcome.
If the coin holder vote is done via QF, without sybil resistance the most likely outcome is that it fails either to provide good information about market sentiment or the views of small coin holders.
To make coinholder voting more useful, I would actually take an approach where registration windows are announced only for periods where the window is already closed, in order to bias the results of coinholder polls toward those who consistently maintain a balance in the shielded pools. We really need good tools like shielded hardware wallets for this to make sense. Of course, to some degree this introduces other risks; if someone can predict when a proposal will be put to a vote, they can speculatively and in advance try to time the registration window. Still, that’s a bigger hurdle to overcome.