Gain staging, how low should you go?

If you don’t have the time to read this wall of text I’m about to type here’s a summary in a sentence or two

Should I compensate for gain/drive within plug-ins or go back to initial vst audio source?

So I just recently upgraded my rig to where I no longer need to bounce but maybe once due to my uad plugs, have 22 sharcs right now but prob need another 8 to achieve a no bounce session, either way i’ve never had the benefit to be able to gain stage so far within a mix as i’be always needed to freeze and flatten due to cpu and never got low enough gain staging to balance everything.

So recently i’ve been doing some tests where I turn down the vst/audio source to a healthy level and start throwing on mixing plug-ins. Let’s say the first one is a neve 1073, I’ll use the gain knob to turn things up +5db, now I’m close to -2db on the track so I’ll go back into the vst or audio source and compensate this way by turning things down there, but everybody i’ve been talking to would tell me to compensate for the gain on the 1073 fader, which way is better? My ears tell me that you into the intial audio source and turning things down there sounds better, also why would i want to keep adding unnecessary steps when I can address the issue from the source. I will then throw on a Studer and start to drive, again now I’m driveing a little hot so I will go back to the vst and lower there as opposed to turning down the output on the Studer, I find that doing things this way allows a much cleaner sound as opposed to doing the up and down thing within the plug-in which makes for a more congested muffled sound imo. Then i got to thinking that maybe by the time I get through all of my gain staging there will be so little sound that it’ll sound thin?

Until I came to one of the reductive eq videos in mixing foundations and saw the stems that Danny had been given. I mean the kick on the pro q was coming in at -12db, and the bass at -20db, and the wave forms were extremely thin, this is exactly the way i’ve experienced gain staging with the best results but haven’t been sure it was the proper way to achieve the best sound or if the pros did this way. So before turning everything into audio I will throw the mixing plug-ins on for a rough estimate also mixing into a Ampex ATR 102 on the master channel watching the headroom, is this the way to do it as far as finding the optimal gain stage level?

If so what about vocals, I know that recording with a decent amount of gain gives crisp saturation while recording lower provides clarity but thinness which later can be addressed with compression and other plug-ins, but should a vocal wav form look similar to the wav forms of the other tracks? Or are vocals a special exception? Not too hot but still heathy?

I know there will be times to turn things down within a plug-in but I just don’t think doing it ritually whenever driveingit harder is the best method, takes away from the juice of what your trying to acheive imo, but countless of producers and engineers say you should leave a plug-in the way you came in(level wise), I say no, it’s betrer to go into the original audio source and do it there, how about you?

Sorry about the run on sentence and punctuation I’m on my phone and in between jobs and just needed to get this out there, thanks