Yes, I got the point. The approach is every bidder doing the zkproof to create a receipt of their bid because the auction will end at some point in the future. I understand it works very well, and it is the most used approach.

I was wondering about the possibility of another approach and my point was creating just the winner zkproof on the auction side (not bidder side), at the end of the auction. If all the bidders used some stealth method plus homomorphic encryption, the bids could be hidden and maybe they could receive the 'winner zkproof bid receipt' and check it against their very own bid. Then the winner could be able to prove it being the owner of the keys who delivered the stealth bid which won the auction.

I read the paper again and understood their approach. It is clear why they are doing the zkproof for every participant in that context.