Welcome to Interact.
For #2, it's better to do all DSP in real-time in my opinion. Store the highest possible source file you can, and apply changes during playback, conversion, etc. If you're a photo guy, this is a bit like using parametric image editing so you always have a perfect copy of the original.
For #1, it's splitting hairs. Both will sound indistinguishable from the original. For fun, you might use DSP Studio > Analyzer and set it to watch those really high frequencies. In most high sample rate material, there's simply no energy at the really high frequencies.