Not sure this is related, but as you know there are two huge boxes—relative to the rest—that (possibly) mess up the distance search. I took them away to test {interim} results. Here I’m pulling points to the {kept} boxes rather than just measuring or testing for inclusion*.
poistclosesttopoint_temp.gh (37.2 KB)
*Regardless of that, you will still find points that are definitely outside of the boxes whose centroids you’re measuring distances to. Have you specified what’s happening with that?

