Render to disk always reverts to the actual playback sample rate, because DSPs are expected to sound slightly different at other rates, and a few plugins do not handle sample rate changes correctly at all.
This way we’re trying to make sure that the rendered result sounds like the song you’ve edited/listened to before…
It does not make sense, sample rate should be remembered as it was set before, maybe only for a session but it’s stupid that priority or interpolation is kept but not sample rate and I have to change it over and over every time I render audio