I should've made this a discussion since I have comments but here they are:

The precursor PR has a lot of context that'll help answer the questions in the PR Review Club page. Particularly I found the comment about tricking peers into thinking the time is earlier than it is (to force them into rejecting a valid block outside their max time range) pretty interesting. It makes me think the attack might have been tried before.

The specific change to prevent the timestamp validation using the adjusted time is pretty straightforward, but there's a lot of junked code from it. Personally I like changes that identify and remove unnecessary code because it makes the overall codebase more legible. I also appreciate the fact this replaces a piece of error correction that may not be correct with a warning message instead. Especially since it's an uncommon error that might indicate suspicious activity.