Re: Launch rail and backplate issues
A vertical launch tube is a nice idea except that you'd have to rig it between at least three balloons for any degree of stability and it would need to be (at a rough guess) at least twenty metres long, to reach from the bottom of the tethers up past the mid-point of the balloons, to ensure that the end of the tube clears the hugely expanded balloons. This tethering rig would have keep the tube centered throughout the considerable expansion range of the balloons. It could be done but it would add a _lot_ of weight, not to mention a folding wing system for the aircraft (which would mean both extra weight and complexity).
There's a simple problem with putting any sort of air-tight cap on the launch tube too. Any cap that's flimsy enough for the aircraft to simply fly through would have burst at a relatively low altitude i.e. < 30,000 ft due to the outside air-pressure drop.
A couple of other people have worried about achieving lift from the wings during the rocket burn phase. This is actually a bit of a reverse issue because if the wings generate any real lift during rocket burn then there's a risk of the plane looping. This is because sub-sonic lift is proportional to airspeed whilst the attitude of the plane, whilst under thrust, is pretty much irrelevant; the lift will always act nearly perpendicularly to the wings so, for example, if the plane were to reach the vertical then any wing-generated lift would tend to pull it back past the vertical to inverted - not good. The wings really don't want to generate too much lift and are really only there for getting back down in a controlled manner after the rocket has burned out. For similar reasons, the launch rail need only be long enough to stabilise the plane whilst the rocket thrust stabilises. Once the rocket is burning steadily that's really the only force that's significant - all other forces will be tiny in comparison.