ESA indicates that the first bounce lasted two hours and reached a height of 1 km. With the extremely weak surface gravity and low escape velocity of the body (< 1 m/s), and other publicly available information regarding the comet's mass properties and the landing site location, can we estimate how close the lander came to drifting away into space?
We can do a very rough back-of-the-envelope approximation, at least. Say 67P is a sphere 1.7 km in radius with mass of 1x1013 kg; Wolfram Alpha says surface gravity is ~2.3x10-4 m/s2 and escape velocity = 0.886 m/s. Gravity 1km above that (at periapsis) would be 9x10-5 m/s2.
I'm too lazy to do the calculus of a ballistic trajectory through that gravitational gradient so let's just use the average, 1.6x10-4 m/s2. If the first bounce is a 2-hour parabola, then periapsis is at 1 hour (3600s), vertical velocity is 0, and by v = v0 + at, initial velocity was therefore 0.576 m/s; periapsis works out pretty close to 1 km with that average acceleration as well, so that seems sane.
Running the same equations for just the small lobe of the comet produces similar results; the lower mass is somewhat cancelled by the lower starting altitude.
So this was a fairly close call - Philae may have taken off at about 2/3 of escape velocity!
It was never close. Rosetta did not have escape velocity when it ejected Philae. So as long as Philae was ejected in the retrograde direction, Philae would not have escape velocity either. After the bounce, Philae would have even less energy due to landing gear attenuation and would therefore be even farther from escaping than it was pre-bounce.
The new report says Philae was moving at 38 cm/s.
Using Russell's figure of 0.886 m/s as escape velocity, we have 43% of escape velocity.