AF447 is an amazing story of a lot of holes in the cheese all lining up…
In your list, you omit the biggest hole of all: Most of this happened while they were flying through and inside a massive CB cloud! Severe turbulence, alternating g forces, lightning, noise and fear/panic take away 80 percent of your mental capacity at once. Then they quickly lost all of their automatic control and found their aircraft in “direct law” where the sidestick controller directly controls the angle of the control surfaces. Something that they had trained briefly during their initial type rating, but which every Airbus pilot calls pretty difficult even in quiet air. So their reaction of cutting the power and and putting the nose up to get as slow as possible was the best they could do! You have a chance to recover from a stall, but not if your wing or tail gets torn off. Why they kept the aircraft in that attitude after the worst was over we will never know.
We too now train for high-altitude upsets with contradicting instrument indications. Even if you know what is going to happen and how to react, it is still quite scary when you get stick shaker and the overspeed warning at the same time and have to decide which one to ignore.
It seems it was ice crystals blocking the pitot. Not airframe icing in the AF case.
In your list, you omit the biggest hole of all: Most of this happened while they were flying through and inside a massive CB cloud!
I would think that is my points 3,4.
I don’t think they were supposed to do that – exactly as “we” are not supposed to do it.
And should one find oneself inside a CB, presumably the procedure is attitude+power (or just attitude, if like me you are usually at max power when above FL100 
), anti ice all on max (if you have any), keep it well below Va, and hang in there for dear life until you come out the other side.
Also you are not likely to have duff attitude info because that comes from an inertial source, and if you lose that as well as fly into a CB, you are having a seriously bad day! Same with altitude. Very unlikely to lose that. But I don’t think the AF447 knew they could trust their attitude – even if they noticed the indications.
And should one find oneself inside a CB, presumably the procedure is attitude+power (or just attitude, if like me you are usually at max power when above FL100 ), anti ice all on max (if you have any), keep it well below Va, and hang in there for dear life until you come out the other side.
I don’t know how it is in an Airbus, but in most planes it is as you describe. Maintain attitude, forget about altitude and heading, don’t chase the speed. We have a “penertration speed” that we are supposed to maintain, unfortunately this is so different from our normal cruising speed at altitude that I would have no idea what power to set in case of losing the airspeed indicators. I think I would start by pulling the levers about halfway back and see what happens… And use the speed brakes rather than putting the nose above the horizon. But I know all that now, before AF447 it was not trained in the simulator.
Quote from a few pages ago:
It is quite scary to realise how little operational knowledge one gained from the flight training machine.
Being part of the “flight training machine”, yes I hear you, but please consider that there is alot more to teach a student than IMC flying in Icing conditions as is being discussed here. For one, it isn’t legal in many typical training aircraft, so that sort of defies the point. If you’re doing twin training that’s another matter, find a Seneca or DA42 and test it to your hearts desire.
It’s all good to think that one should get to fly with the “experienced” real world IFR GA guy who is the best at what he does and all other instructors are crap, but then, why aren’t all you experts here instructors providing this service to the new pilots?
My advice to “new pilots” is to get the basic training under your belt, fly in good weather (no ice) with your basic aircraft, bring an experienced friend along and pick his/her brain then move on up and apply the same procedure as you go. That’s how it works in the airlines (although somewhat accelerated)…
On a more personal note with regards to Ice flying in non-deiced aircraft (I have done it, yes). It is plain stupid, and I was a fool.
Escape plans are great, and the best one is; don’t fly. Who ever really needs to fly?
In your list, you omit the biggest hole of all: Most of this happened while they were flying through and inside a massive CB cloud! Severe turbulence, alternating g forces, lightning, noise and fear/panic take away 80 percent of your mental capacity at once.
Nobody on this thread seems to like reading accident reports:
the AP disconnected while the airplane was flying at the upper limit of a slightly turbulent cloud layer
In fact the turbulence amplitude never exceeded 0.5 g peak to peak, light by any standards.
A constant headwind component of 15 kt had to be added to make the simulation’s ground
speed match the parameter recorded. This value is consistent with the wind parameters
recorded. The results obtained reveal that before approximately 2 h 10 min 40, i.e. the time
when the aircraft was climbing at about 37,000 ft, the parameters recorded (angle of attack,
normal load factor, and attitude) fluctuated around the simulated parameters, indicating the
presence of turbulence. After this time, this turbulence appears to disappear and the
parameters simulated and recorded are highly consistent.
Consequently, it would appear at this stage in the work that the bulk of the aircraft
movements in the longitudinal axis (attitude, vertical speed, altitude) result from the actions
of the PF, with the exception of small variations that are probably due to the meteorological
disturbances.
An on topic article by Richard Collins
http://airfactsjournal.com/2013/12/ice-gotcha-heartbeat/
Here some pictures.
Deicing the aircraft in the morning.
After that I was ready to fly.
Then once above this thin freezing fog layer, I got this view.
Ice can be dangerous, but it doesn’t mean you can never fly.
I had this idea that knowing the amount of LIQUID water in a cloud would mathematically provide an estimate of worst-case-icing “survival time” in that cloud.
From this article
http://www.das.uwyo.edu/~geerts/cwx/notes/chap08/moist_cloud.html
one knows that even the darkest of the clouds cannot have more than 3grams/cubicmeter of liquid (possibly supercooled) water. A plane frontal section, excluding the prop, is around 1 squaredmeter, so, at 130kts it will “touch” 4000 cubicmeters of cloud every minute. Assuming all the liquid water in this space sticks to the plane as ice, this would make 12kg of ice every minute, i.e. 1.5cm of ice thickness per minute assuming again a frontal section of the leading hedge of wings and stabilator of 1squaredmeter.
I would say this result is in line with the worst experiences various members have reported in this thread, and it provides a maximum theoretical worst case (unless there is freezing rain, which is another story).
