VocalRider is kind of expensive. If you were in Reaper, I know somebody wrote a script that does about the same thing and apply it directly to the take volume (what PT calls clip gain) envelope. But really those methods are just RMS compression, maybe with some lookahead, and that's how I do it.
Find a compressor with a "lookahead" or "precomp" parameter and an RMS parameter. Set RMS pretty long and lookahead exactly half of that. I use ReaComp, and its precomp parameter tops out at 250ms, so I set RMS to 500, and it works fine on most material. Attack and Release to 0 (or as low as possible) and then ratio doesn't have to be so big. On vocals I sometimes end up as high as 4:1, but more often down around 2. You kind of have to set that to taste. It works best if you set the threshold so that it is almost always doing a little bit, and even better if it's got a soft knee.
A lot of people will tell you that doing it by hand is the "right way" and everything else is cheating, but when it just works almost all of the time and saves me time and frustration so I can do other, more important things... I'll sometimes manually adjust things that are really way off, but then I'll gladly cheat.
But does PT not have a clip gain envelope? It must, if you're talking about recording a MIDI knob. I would think that drawing curves there would be at least a little faster than splitting it up and adjusting each chunk separately.
Yes it probably should have been handled in tracking, and yes you might be too caught up on numbers, but frankly this kind of leveling makes almost every vocal sound a little bit more polished and ready. It can make a good performance better, and a great performance...well...better.