So I have a vocal bus where the lead, bk and anything else muttered goes to.
I did a volume envelope on it to ride the vox thought the loud and soft. (I know about vocal riders)
If I want to bring up all the vocals by say 2db though out the song, because of the envelope's "ownership" of the bus volume I historically just raise the output gain of the last plug in the vocal bus. The bus input gain is available but is just swallowed up by any dynamics plug on that bus.
 
How should I really approach this?

You should go through each track and level out the volumes on each track as pleases you.  You should also put ~3db compression on every one of those vocal tracks.
 
You should then mix and route all those vocal tracks to an "Aux Vocal track", and the aux track to a stereo outbut bus (labeled Vocal), which in turn is routed to the "Master" output bus which is routed to your soundcard.
 
When I get all the vocal tracks' volumes as best I can, I RECORD that whole mess to the Aux track, and then mute/archive all the individual vocals.  You can then automate the volume/pan of the Aux track vocal mix. [This also facilitates taking advantage of the "vocal fattening" technique - if interested send me an email and I will provide the procedure.]
 
If you want the overall vocals louder/softer, simply adjust the Vocal output fader.
 
Best results are obtained via compression by chaining so no one compressor has to do all the work .... at the vocal track stage, at the Aux track stage, and on the Vocal bus ... if you need to, if you want to, if whatever......

∞

Ride the actual vox track?  It does get hard and confusing, a rubix cube of sound.  Another technique is to put all the music through, well a music bus.  I do that with some spoken word songs I'm working on.  I can lower the music when it starts impinging on the clarity of the Voice.  Just use an envelope for the music and all your vocals will stay balanced as before.  
 

bitman
So I have a vocal bus where the lead, bk and anything else muttered goes to.
I did a volume envelope on it to ride the vox thought the loud and soft. (I know about vocal riders)
If I want to bring up all the vocals by say 2db though out the song, because of the envelope's "ownership" of the bus volume I historically just raise the output gain of the last plug in the vocal bus. The bus input gain is available but is just swallowed up by any dynamics plug on that bus.
 
How should I really approach this?

You are over complicating things.
Do not use Vocal Rider(s), avoid fixing things in the mix, instead get it right by doing a great take.

This is what set professionals apart from amateurs. Take your time, get it right.
Sometimes I'll spend a month recording a single part (I have the luxury).

Here is what I do:

1. DO NOT place all of your vocals on a single bus. Instead break it into sections like this.
   Verses --> BUS -- > Master
   Chorus --> BUS -- Master
   Adlib   --> BUS -- > Master
2. Apply processing. E.g: compression to each vocal bus as desire.

Yes you can ride the bus but I prefer to get a consistent take.
For my latest song, I wanted the chorus to stay back in the mix and to have more Stereo Room.
This would not have been possible if all of the vocals were going through a single BUS, and I rather not process individual vocal track.

Do the same with instruments, especially Guitars or anything requiring dynamic processing.

In the end nothing I or anyone say means anything, do your music the way you feel makes sense!

AI

Original Pranksta
Offset Mode

 
Thanks you!

Obviously there are hundreds of way to approach this and we each have our own technique. There is never one correct way to do anything with all the track,  buss and insert options we have available in a DAW. 
You are at the stage of seeking which method will work for you.   
Explore the options and decide which one works best for you. 
 
I'll add one more method for you to ponder. 
 
First I try to record the lead vocals nice and hot buy using a hardware compressor on the way in. It is not used to compress, just catch peaks. 
I then add a PC 76 compressor to the tracks Pro Channel I just use the default setting  
If there were any obviously low level parts in the track I add gain to even it out. 
And automation can certainly be applied here if needed. 
The vocal tracks are then all sent to the Vocal buss where I use a second compressor but I turn the WET/DRY to around 1 o clock which I think of as being the same as using parallel compression. 
This has made it very easy to have in your face vocals and using the buss level at any desired volume in the mix. 
 

I also use offset mode for something like what you're talking about. I used to use buses before I knew about offset mode. 
 
My gosh, I can't imagine not using a vocal part that was otherwise great just because it didn't match the volume level of the other parts. That's what automation is for. I have better things to do with my time. Professionals definitely ride vocal buses and/or use automation. 

You can also select the whole automation envelope on the Bus and raise and lower it all at once.

Hi.
Offset mode is a favorite of mine as I always have the need to do exactly what you're doing, due to adding stuff towards the end.
All the best.