I get where you're coming from. This is another argument I've heard before. The question I'm trying to answer isn't whether one can use 16-bit instead of 24-bit, or one would experience "less noise" in 24-bit, it's what is happening within the electronics and data spaces.

In photography, I shoot RAW because I want the things you mention. But if I expose where I'll ultimately want things, JPG (8-bit) is fine. RAW isn't magic to me. It's just more data. I could talk about it the way you do--I probably do--but I don't want to romanticize what's going on. There's no magic.

A few bits of 24-bit are "generally" not anything, noise or otherwise. Reconsider?

I understand the benefits of "float" data in computer work. I've made much of my living with Excel ;) My focus here is recording only and as I argued 32-bit float is 24-bit precision and if fidelity is your goal you don't want composite (exponential number pairs).

Again, let me know what I'm missing! Thanks for comment.