What you quote is for conversion with analog levels, which is not what’s happening here. Everything I’m doing is 24 bit digital audio, which has 144 dB of dynamic range. That is a little over-kill, which is why most audio files are distributed as 16 bit, so 96 dB. That means you can hear anything from 0 dBFS to -96 dBFS (with proper dithering). That is why the cutoff point in the graphs I showed you is -100 dBFS, since you realistically won’t be able to hear below that anyway (audiophiles disagree), in the final file.

-40 dBFS only represents how the signal is stored in the digital file. It has nothing to do with the signal’s actual volume. I play those -40 dbFS through my computer, then my DAC, which outputs at about 2 VRMS, into my pre-amp, which increases the voltage again, into my amp (which, again, increases the volume), and finally into my speakers, which output that -40 dBFS, which is now signal at about 70 dB SPL (actual volume).

dBU is for analog line-level, and the conversion you showed is what I would use when routing my console back into my computer. But analog line levels are still very audible at -40 dBU, usually, not that that’s relevant.

You have to understand that this is not real volume. It is just how the volume is stored digitally. If you have a -40 dBFS noise floor in your audio file, and the music has a 12 dB dynamic range peaking at -1 dBFS, you will hear the noise clearly throughout the entire track, because you are amplifying the entire thing greatly, since you are ultimately transforming this into dB SPL.