Consider optimizing `RightSemi` to eliminate dups from build side

### Is your feature request related to a problem or challenge?

#22914 means that we will no longer produce intermediate duplicate output rows for `RightSemi`, but we still store duplicate build-side rows in the hash join build side. We could consider eliminating that. This would reduce hash join memory consumption, but the tradeoff is that we might do some wasted work if we spend time eliminating dups that would never participate in the join in the first place. Merits some further study; we could perhaps make this conditional on the fraction of duplicate hash values we observe on the build side as we execute the operator.

### Describe the solution you'd like

_No response_

### Describe alternatives you've considered

_No response_

### Additional context

_No response_