Overload some floating point functions for integer types

Functions such as `isinf` which do not rely on integer->floating point type promotion could be overloaded for non-floating point number types.

Start by locating the Multiprecision implementation and overloads for `isinf`, then identify the other floating-point functions covered by the same proposal. Determine which functions do not require integer-to-floating-point promotion; done means those functions accept non-floating-point number types with appropriate coverage.

Written by the indexing model from the issue text.