Having a high angle of attack can come at any speed. It's all got to do with relative airflow. An aircraft can be descending and still have a high angle of attack. So MCAS has just got to do with getting the nose down and below that high threshold. The high speed means it's available throughout the entire flight envelope.
I think we need to separate some of our terms here. And I'll start by saying that I'm still concerned about MCAS, and don't see why it is needed outside of a fairly limited range. It is a form of angle of attack protection, but because Boeing have basically 'cheaped out' it doesn't allow the degree of fine control achieved by a proper FBW, in which you can fly the aircraft right on the AoA limit very accurately. I'd be interested in seeing its behaviour in a GPWS/terrain event.
Angle of attack (alpha) - the angle between the wing and the relative airflow.
Rigging angle - the angle between the wing and the fuselage. It's what caused aircraft to fly with the cabin very slightly nose up in the cruise. In large part its chosen for approach and on ground behaviour reasons.
Indicated airspeed (IAS) - the airflow you'd feel if you stuck your hand out the window. Variations called 'EAS' and 'CAS'.
True airspeed (TAS) - IAS corrected for the difference in the atmosphere. In simply terms, if the air is half the density, you'll need to go twice as fast to show the same IAS. The upshot is that at sea level IAS and TAS are basically the same, but as the altitude increases, the TAS for a given IAS also increases.
mach number - aircaft TAS expressed as a percentage of the local speed of sound. This is limiting at the upper end because various nasty effects start coming into play above the .85 to .9 speed range. In practical terms, mach number provided an upper limit on speed at high altitudes, but is largely irrelevant at lower levels.
So, how do you have a "high angle of attack at any speed". Well at low level you've got two options. If you're trying to maintain level flight, and you slow down, your AoA will increase. Keep slowing, and you'll eventually reach the stall AoA. Alternatively, you can apply some roll, and start to turn the aircraft. To maintain level you'll need to increase the g loading (as part of it will now be dedicated to making you turn, you'll need more overall to counter the 1 g of normal level flight). You increase the g load by pulling more and more backstick, which increases the AoA. Stall speed (which is associated with a more or less fixed AoA) increases. At 2g (a 60º angle of bank turn), it will increase by 41%. At 75º, you'll be generating 4g, and the stall speed will have doubled. So, if your 737 has a clean stall speed of 180 knots, it would have gone up to 360 knots at 4g.
In the cruise (at, say FL370) , you're limited by mach number. If you have a cruise speed of .83, that will give you an IAS of about 270 knots. Your TAS will be about 480 knots. If you apply the wind to your true heading and TAS, you'll end up with the track and ground speed. That's what gets you from place to place, but doesn't actually make the aircraft fly.
Now, at our .83 cruise at 270 kias, we're about 90 knots above the clean stall speed. But another number that comes into play is the minimum drag speed, and as it will be around 250 knots, we are only slightly faster than it. The aircraft won't immediately do anything nasty if we slow below min drag, but the issue is that at high altitude you won't have sufficient power to accelerate above it again, so the aircraft will start to slow down, and even with maximum power applied will continue to do so. If you're foolish enough to keep it up, you'll eventually end up with stalling becoming a consideration. The only solution is to trade height for speed.
So, lets put turning back into the equation. If, in the cruise, you roll into a 60º of bank turn, we'll need to generate that 2 G to maintain the altitude. Min drag speed, and stall speed will both increase by that 41%, so we're now well below min drag (now around 350 kias) and our stall speed has now jumped to about 255 kias. As we're so far below min drag, the aircraft will quite rapidly decelerate back onto the stall. Yuk. Because of the effects of turning on these numbers, in airliners are generally limited to 30º, and the autopilot will restrict its own turns at high level to below 20º.
The upshot of all of that is that the stalling AoA is never all that far away, and it doesn't really take much to get there, so some way of protecting the flight envelope becomes desirable. In the past this was up to the pilot, and then we had various forms of assistance with stick shakers and/or pushers, and eventually the ultimate form of protection in the form of FBW. But, there can be a reason for flying back near the stalling AoA, for that's where the most lift is produced. For example, perhaps an aircraft on approach runs into a microburst. Not only will it need maximum power to get out of it, but it will need to be flown right back at maximum AoA until it is in the clear. In something like the 747, you'd pull back until the stick shaker activated, then relax the pressure just enough to stop it. Holding exactly on the point of activation isn't easy. In a FBW aircraft you pull full backstick (all the way to the stop) and hold it there. The FBW system will continuously and aggressively activate the elevator to hold it on the cusp of stalling (they'll have picked an exact angle to use in the development). But the FBW systems never use the trim system for this. I don't know how MCAS would respond in this situation (AV?) but it is a case where any system that reduces your ability to get more AoA would seem to be a negative.
Stalling. There are two types of stalls that are mentioned in aviation.
An jet engine compressor stalls when the airflow through it is disrupted. This can be caused by damage from a birdstrike. Pulses of flame and banging are associated with this. The engine is not on fire. Generally the engine will be shut down. They can also be caused by airflow disturbances from extreme manoeuvering. Not normally an airliner problem, but the engine will normally recover by itself if the power is reduced to idle, and then advanced once normal flight has been resumed. Airliner engines in reverse can be prone to this, and it generally self recovers once selected out of reverse. They can also stall during start, and this will lead to an aborted start. You don't stall engines for fun, as it can be very damaging.
An aerodynamic stall happens with the airflow over the wing breaks away, and the flow becomes turbulent. At that point the wing mostly stops making lift, but does make lots of drag. It happens at about 15º angle of attack. Remember that the angle of attack is not the pitch attitude of the aircraft. Whilst stalling airliners isn't a great idea, it's a basic skill that's taught to new pilots, and many of the aerobatic manoeuvers you see at airshows involve going well past the stall.
Some aircraft have a particularly nasty charactistic in which the turbulent airflow from the stalled wing goes back over the tail plane, and diminishes its ability to give the nose down input needed to recover from the stall. Aircraft with T-tails are prominent here.