Many aircraft have noticeable buffet at stall...but many do not...
One of the reasons that there is aerodynamic twist in wings (change in the mean chord angle along the length of the wing) is so that the wing root stalls first and allows the wing tips (and therefor, ailerons) to remain unstalled when that buffet happens so that roll control is still available during the early parts of the stall.
A wing will always stall at the same AOA - so, depending on load factor ("G"), density altitude, etc. the airspeed at which that occurs might be very, very different.
In airliners, and in Cessnas, pilots never really fly into deep stall, they get to buffet, in which part of the wing has stalled, and then fly out. In a lower powered airplane (Cessna) this requires lowering the nose in addition to power. In a better powered airplane (e.g. Q-400) you can simply power out.
But in a truly deep stall, all the power in the world can't get you out...the F-14, as an example, did not have a well-defined stall...at 10 units (close enough to call them degrees), the wing would buffet, just a bit, but you could still fly the airplane (with spoiler and differential tailplane for roll control)...at 15 units, the stall was worse, but you could still roll and power out. 15 Degrees AOA, by the way, is well above where most airplanes stall.
When you got up to 25 degrees, the buffet was similar, the power required was far greater and roll control was degraded.
But the big tailplanes of the Tomcat, which were sized to handle supersonic trim changes (another topic for another day) could take the jet well past 30 degrees AOA. Pull back on the stick and you could peg the AOA at 30 (it was closer to 60 actual) and the airspeed would drop to zero. You would then be well and fully stalled. Select max Afterburner, and all 50,000+ lbs of thrust would not break the stall (that's roughly 100,000 HP, but who's counting)...the plane would descend, in full AB, with zero airspeed, at over 10,000 FPM...
You had to fly the jet out, breaking (lowering) the AOA to allow the wings (and fuselage to some extent, but again, another story)to fly again - and the power would help, but it took a change in AOA to break the stall.
Both the Colgan flight and the AF 447 crew were in deep stall through pilot action - pulling back on the yoke/stick. In the AF case, that was the trained procedure, and it worked when the jet was in normal law, which limited AOA. However, AF 447 was in alternate law due to the airspeed sensor failures, so the pilot was not commanding optimum AOA with full back stick, as he thought he was, he was commanding max elevator deflection, and therefor, max AOA, so the jet never recovered. In the case of Colgan, the pilot pulled back to 20 degrees nose high, well beyond stall AOA, even with flaps down, and the roll control went away as the entire wing stalled...and all the power was useless in recovery from a deep stall...
