Sorry I'm a bit late checking into this thread.
My first shuttle full stack was 1/110 scale. Orbiter was sort of crude balsa, but was very light and glided nicely (I have rarely seen the Estes 1/162 shuttle's orbiter glide at all). To me, the 1/162 kit was just too small.
I feel 1/110 is close to optimal for flight, with BT-55 SRB's and 3" ET, as at that size and weight (if built with weight in mind), it will fly nicely on a D12.
I know 1/100 is preferred in comparison with other 1/100 kits, to stand all together. But for models that FLY, I don't think the scale factor compared to any other model is as important for optimal flying and practical factors of assembly and cost and reliability.
In 2009, I made another 1/110 model. This time, with the Guillow's foam orbiter, which I'd used as a piggyback orbiter on sport boosters before. It flew well too, on D12 power and also E9.
It was a prototype for a kit, but I never ended up producing it. Some more info:
http://georgesrockets.com/GRP/Scale/Shuttle2009.htm
Shuttles at 1/72 size end up so big and heavy and draggy that they pretty much need a G motor. Although my 1/72 shuttle stacks flew on F motors, mostly the original Aerotech F25 (close to 80 N-sec). But those were made lighter than some other models would be. I mean, the NCR 1/72 shuttle absolutely needed a G40 to fly, due to its' mass.
My 1979 contest shuttle, which won scale at NARAM and 10th at the 2002 World Championships.
I don't have any good video of that shuttle. But here's a test flight of the 1998 boilerplate after many years of trying to get a shuttle to work with SRB's that sepped and so forth. Warning - the sound is loud.
https://www.youtube.com/watch?v=OQbt2VoIuZo
"fun fact" - the orbiter was 14 years old then, built in 1984. Used a vac-formed orbiter nose grafted to the balsa orbiter. My contest model was built the same way. Note the steering was by the lower part of the rudder, not mixed elevons (elevons were used for elevator only on my model). Orbiters can be steered beautifully using the lower part of the rudder, mixed ailerons do not fly smooth.
When it worked 5 times in one day, and 3 times another day, within a span of about a month, I knew I finally had all the problems solved. After so many other boilerplates and even trying different scales, and finding out clustering was not worth screwing with. Also part of it was methods I had learned thru those years, plus technology that did not exist in the early 1980's. The ET nose has a flight computer in it, made and programmed by Jay Marsh, in control of a lot of things (such as SRB sep after burnout was detected, and ET ejection charges) except for the orbiter sep which was manually by R/C.
And, here's a
link for the shuttle pages on my website. Two involve my models, the last is a trove of shuttle data.
http://georgesrockets.com/GRP/Scale/SHUTTLE.htm
Last thing, a video by Greg Warren showing my 1/60 scale obiter using a very special piggyback booster, powered by a G12. Glides nicely. It also used rudder for steering, so it turns smooth like the real thing.
https://www.youtube.com/watch?v=e_cxNTnIYsA