[iwwr]

The 100km is not arbitrary. It's an altitude past which a wing would have to travel faster than the orbital velocity in order to retain lift. It's a model that only considers angle of attack and the density of the atmosphere. In other words, it's the maximum height at which an aerodynamic vehicle can retain altitude (where lift >= drag).

[lutorm]

That makes no sense. Looking at that Wikipedia article, it says lift L=0.5 rho v^2 A C_l. For flight, L = mg. So for a given speed you get a criterion on the density of the air:

rho = 2g/(v^2 C_l) * (m/A)

So the density at which you can fly depends on the surface density of the structure. Which of course makes perfect sense -- if you build a lighter airplane, it needs less lift.

To get that altitude, Karman must have used some typical surface density of an ordinary airplane at the time. Presumably we can now build much lighter structures, so we should be able to fly higher.