SE-0259: Approximate Equality for Floating Point

I concur. Especially if (a-b).isAlmostZero(absoluteTolerance: ...) is a naive and wrong way to do it (it's easy to grow a small amount of paranoia when dealing with floats).