Just to confirm - "Whoever stakes the most GTC, ranks the highest" == "Whichever grant has the most GTC staked on it, ranks the highest" - correct?

correct!

In 6. - Are we saying by burning GTC to downvote a grant, I also burn someone else's GTC who earlier upvoted that grant?

Is there any incentive for the staker to take these actions (beyond trying to elevate a grant in the ranking?). ie, do they earn a yield by doing so (perhaps an additive % of the funding the grant receives during each funding period?) Im wondering if ranking changes will be sufficient on their own to encourage participation. further, the risk feels large to lose GTC should someone with more GTC decide to "wipe me out."

No, and I removed the phrasing to be sure. What actually happens is that the GTC I pay to downvote a grant is sent back to the staker, while the GTC removed from the stake is burnt. The cost to downvote is always less than the effect on the ranking. I think ranking changes should be sufficient, but this is an opinion that really needs to be tested. It may be that we need to encourage further participation, which may take the form of turning inflation on and using that to allocate to stakers based on how much they have at stake. My suggestion is to go with the simple version first and only add rewards if they prove to be necessary. Appearing in the first page of grants is, imo, a very strong incentive in and of itself.