Thanks for the replies!
The simple fact is: If a human can solo each track and export it, with the inclusion of any effects or buss processing, then a computer can do it; and in one go. It's just an oversight in the export features.
 
Obviously if one has many different things going to a buss with processing on it, the processing might act a little differently when each track going to the bus is soloed; but it's very close; and many artist, film and tv projects expect stems that sound like their stereo mix when played all together, so that they can make small adjustments if necessary, after-the-fact and re-export.
 
Further to this, I think there should be a way to deliver stems where even when the tracks are soloed, the buss processing acts exactly as it would with all the tracks enabled. This would ensure the same mix once it's in stem format. A possible method of doing this might be to give each track an option to send its signal to the bus in prefader mode such as it would if you set up auxiliaries; so if you turn the fader down, the track still sends to the bus processing chain but is not audible. (like when you set up a side chain as a ducker). I may be off my rocker but I think the basic concept of this could evolve into a more accurate way to mix-down stems.
 
JG