Why doesn't the API.Board use the same draw-by-repetition detection that Chess.Board does?

[ginop]

I would like my bot to accidentally draw less often, but it is limited by the not-quite-right draw detection used for API.Board. Why does it work that way when the Chess.Board has the correct method already?

As an extension of the same question: why does the board's RepetitionPositionHistory not get updated for moves made while searching? If it did, true draw detection would be trivial to implement. Quite a lot of work has gone into avoiding that though, so I imagine there is some good reason that I just haven't seen yet.