If you're getting complete rearchitecture feedback in a code review, something's wrong because that's too late for it. Especially if it means rewriting later changes they haven't put up yet.

The wrong part happened when sender went to work on something alone for a long time then sent a giant PR all at ones instead of progressive set of changes. Don’t blame the reviewer for having to either compromise their integrity by letting it slide into the codebase or getting onto the likes of TP’s shitlist by putting their foot down.