I'll mostly repeat what I said in the beta SO for Teams site.
I don't think this should reward rep. Precedent is relatively weak, and robo-reviewers are already enough of a problem in review queues where all they can get is badges.
If it must be rewarded rep, the proposed values are too lucrative. I propose maximum absolute values of 2, except that publishing spam or R/A should be a review ban and something harsher like -50 rep.
From an abuse-case perspective, you are essentially creating a proxy channel to perform serial-voting on reviewers (assuming SG reviews are or eventually visible in a user's profile page- which they should be). Ex. say I want to act on my pettiness and have something against reviewer A. I can downvote everything reviewer A approves and make their rep go down. I don't see what's so difficult about figuring out what kind of mod tooling you need to detect rep-based abuse- at least at the first order of the most obvious abuse-case this will enable. Integrate the review system into the existing serial-vote-detection system. As in- in that system, treat a vote by user B on post approved by user A as a vote by B on a post by A.
Rep gained is locked in after 14 days to protect against long-tail vandalism.
How is this a net improvement? You want to know what thinking "long game" looks like? "I can make a voting sock, very slowly upvote posts reviewed by my other account, wait for 14 days to pass since each vote, edit the post, and then remove the sock upvote to cover my tracks. Receiver account keeps the rep, but the vote is gone.". Unless mod tooling keeps tracks of votes as transactions in addition to the latest existence of a vote by a user on a post, and factors that into fraud detection, and has parameters wide enough to catch it, that's going under the radar.
We are trying something new, and this is just our first attempt at dipping our toe in the waters of rewarding curators with rep, a form of recognition which we believe is overdue
Why this specifically, then? See for example, https://meta.stackexchange.com/a/387398/997587.
In future iterations, we are open to the possibility of allowing multiple reviewers to gain rep from the same Staging Ground post, rather than just the final approver, as this is often a collaborative process. - Sasha
I can get the motivation (I actually mostly look at just new stuff in SG), but it's a nested can of worms in what's already a can of worms. I can imagine people abusing it in similar ways to how you can abuse "leaving a review" in First Questions by just upvoting a comment. Except now they are doing it to get rep and not just a badge. And what if I only review new stuff, and other people publish something in a state that I don't think is ready? Would you penalize someone for a downvoted post in that scenario? It's messy.