You need to think about signal flow and where you want the level changes to occur. For example using MIDI to set the Volume is basically changing the level coming directly out of the VSTi. Where if you use the MixConsole’s fader the level gets adjusted after some or all of the Channel’s Audio Inserts.
A) Sound Source > level adjust > compressor > EQ > …
B) Sound Source > compressor > EQ > level adjust > …
Both A & B are valid to do, but they will sound different. Neither one is the “right” way to do it although B is likely more commonly used.
Unless you are at risk of running out of computing resources save yourself a bunch of work and go with one Instrument Track for each VSTi. And if you are running out of capacity the solution is gonna be upgrading not trying to crank out some marginal capacity by rearranging your mix of VSTi’s on Instrument/MIDI Tracks (see: deck chairs, Titanic). This will only produce marginal differences and not always in the way you’d think. Search the forum for several long threads on the topic.
Most folks doing huge orchestral templates end up using Vienna Ensemble Pro to host their VSTi’s.