Yes, it is a bit OG -- indeed on a first read of the question I would have probably said "yes" rather quickly and moved on without a second thought, and so been wrong. I suppose the point though is that while, at the instant before the impact, two cars doing 50mph in opposite directions is equivalent to one car doing 100mph, what matters once the impact begins is (kinetic) energy, rather than speed. The kinetic energy involved scales as the square of speed, and 100^2/(50^2+50^2) = 2, ie a 100mph car hitting a stationary car has twice as much kinetic energy as two 50mph cars hitting each other.
So a slightly better (if still probably naive) argument would suggest that the 100mph collision is about twice as destructive as the 2*50mph collision, and further that you'd have to have two cars travelling at about 70 mph head-on in order to try and replicate the effects of a 100mph-stationary collision.
This probably isn't quite right either, mind, but the counter-intuition at any rate is emerging from accidentally assuming that things always scale linearly.