Traditionally, downvotes have proven to be a useful tool. I can think of a few (maybe still around, but in my vybrain) failed projects that harnessed the mechanic of no downvotes. Whaleshares and Blurt come to mind
Yes, removing DVs from Layer 1 without having robust alternative means to combat abuse does not seem viable (e.g. Whaleshares and Blurt). That's why we favor a Layer 2 solution. We can rely on Layer 1 for account-level censorship-resistance while managing abuse at the tribe level. We will use the blunt instrument of the tribe-wide mute for egregious offenses, but we eventually hope to develop community-driven standards and metrics so that every upvote can be a far-more-informed upvote, thus curbing certain forms of abuse, such as circle-voting, at the individual upvote level. For example, the front-end displays each accountholders circle-voting metric at the top of each post, so that before I cast an upvote I can consciously choose to not vote or to lower my upvote if that # is high.