I think you're claiming that it is impossible for two colliding objects to release fragments that have more energy than either of the original objects. That's wrong.
If it helps, start by imagining two satellite-like machines colliding in interstellar space (i.e. not in orbit), where they were initially moving at 1m/s in opposite directions. Even ignoring the possibility of an explosion (unspent fuel, pressurized areas, I dunno), it's still very easy for interactions to violently fling individual parts outward at speeds higher than 1m/s.
Of course, my guess is that the possible amounts of additional delta-v are pretty low, like 100 m/s, and as such the resulting orbit would not be much higher than the original orbit. But that's just wild conjecture.
If you were to collide two LEO satellites and turn them into fragments, even if you give those fragments much higher velocity than the initial satellites, you _cannot_ launch them into a higher orbit.
Fragments that are launched down obviously hit the Earth. But an orbit is a closed ellipse, so the fragments that are launched up will _also_ hit the Earth -- they'll just go up steeply, turn around, and come down. The only fragments that won't initially hit the Earth are the ones that are ejected tangentially, parallel to the Earth's surface. Those will go into an elliptical orbit with a high apogee and perigee at the altitude of the collision. Which means they will _still _come down to that altitude and gradually lose energy.
I agree that the best case (uh, best for people who like Kessler cascades?) is that the new fragment has a perigee equal to the location of the collision, and a higher apogee. That means, yes, that it will suffer whatever degree of drag was present at the altitude of the collision.
However.
The grandparent post claimed that "Fragments that are ejected from the collision cannot have more than energy than they had originally, nor can they change direction very much", which I think is clearly what I addressed in my comment, and I stand by my correction of that, and your reply doesn't address this at all.
Also, with respect to "you _cannot_ launch them into a higher orbit", you just agreed that you can have a higher apogee than the collision. If the input satellites had perigees lower than the altitude of collision, you can also have a lower perigee. More velocity at this point on the orbit, higher apogee, higher perigee.... So how do you figure you cannot have a higher orbit?
Parts could be squeezed between high pressure zones and get shot out of the collision, but even those parts won't gain a higher orbit because the perigee will be below the original orbit.
First of all, no, the perigee need not be "below the original orbit", where did you get that? The correct thing to say is that the new orbit's perigee cannot be any lower than the location of the collision.
But more generally, how does this justify your claim that "Fragments that are ejected from the collision cannot have more than energy than they had originally"? Or were you conceding that point and making a different point?
If it helps, start by imagining two satellite-like machines colliding in interstellar space (i.e. not in orbit), where they were initially moving at 1m/s in opposite directions. Even ignoring the possibility of an explosion (unspent fuel, pressurized areas, I dunno), it's still very easy for interactions to violently fling individual parts outward at speeds higher than 1m/s.
Of course, my guess is that the possible amounts of additional delta-v are pretty low, like 100 m/s, and as such the resulting orbit would not be much higher than the original orbit. But that's just wild conjecture.