Interesting one, thanks for linking it, I missed it.

I used to see the fact that zaps can be faked as a serious limitation to the usage of zaps for things like recommendation and curation algorithms, but as the footnote hints it might not be such a big deal. Even if we can't be sure a payment occurred, we know there was intent from a specific npub to tip a certain amount to a certain content. This intent alone might be a strong enough signal on its own.