Benefit of recording 32-bit audio?

The Sample Rate and Bit size is independent on each other.

The rounding errors are on the Bit side of things, not the Sample Rate of things. Just saying.