My actual question here is not to understand how a wing creates a
lift, but to understand how this is possible: picture of the Venturi
TL;DR when something moves through the air mass it causes the random motion of the air molecules to be interrupted and that causes a different distribution of velocities in the air. At the macroscopic level, we interpret that distribution as pressure.
What follows is a toy model, but outlines the basic concept of "what's going on".
Let's start by considering what gas "really is", a mixture of atoms and molecules moving about randomly. Quite quickly too, the thermal velocity of room temperature air is about 1000 m/s, or 2200 mph! It is also important to remember that their speeds are distributed according to the Maxwell-Boltzman distribution, so there's some slow-moving stuff, some fast-moving stuff, and a whole bunch around 1000.
So consider a little bit of air for a moment. If you could take a snapshot and see all the molecules and where they were going, you could plot them in 3D with their angle from the origin being their line of motion and the distance being their speed. The result would be a sort of fuzzy ball. Note this is not a graph of their locations, it's a graph of their velocities. Do you see that? It's kind of key to what follows.
Now you have all these particles whizzing about, but the net speed is zero. Do you see why that is? In any given volume of air, for every particle at 1,1,1 on your graph, there's one at -1,-1,-1, going the opposite direction and angle. When you add the two, they cancel out. The entire bit of air does this, which is why you don't feel a 2200 mph wind all the time.
Now let's add an open tube to our system. Consider that tiny bit of air you plotted sitting at one end. Which particles on your graph can enter the tube? Well if the tube is to the right and runs off to the right, it's any of the ones who's vector is somewhere along the X axis. The ones up on Z, for instance, are flying off in the wrong direction.
Ok now instead of a complete venturi, consider just the inlet. It's basically a cone:
So now consider just the front part (on the left) and we're going to seal off its outlet (around the middle in this image) with a piece of paper. Ok, now take that bit of air we were considering earlier, and put it inside the cone. Now what happens when you take the piece of paper away? Well, for that instant at least, the only molecules that can get through the new opening are the ones travelling in that direction.
So then if we consider just those ones, they aren't evenly distributed in velocity, you're selecting just the ones moving to the right very quickly. So the average speed of the air in the tube is very fast indeed!
This illustration leaves out an important point - there's already air on the other side of the cone when we open the hole, so it can't really flow like this because the molecules going the opposite direction on the other side of the paper are going into the tube from the back as fast as our slug of air is going the other way.
But that's not the case when you push air into the "front" cone, either by increased pressure or by moving the aircraft you're attached to. In that case the "back pressure" is lower. What is pressure? The number and energy of impacts from the molecules. So if we set the number of particles to be the same on either side, which is a good bet in the case of a plane for instance, then the difference is due to the speed distribution. So the distribution of velocities on the front are different than on the back, which means there's some tiny leftover net velocity towards the rear. And if you consider the fact that you've selected those particles travelling along the tube, you can immediately see their trajectories result in less impacts on the side of the tube, which means lower pressure inside the tube.
I said this is a toy model. That's because air doesn't really move like this, in most cases any one particle gets maybe a couple of nanometers before bumping into another molecule and quickly randomizing everything. So it's not like you have these "left-going molecules" getting all the way through the pipe, they're barely inside the entrance before they're all over the place again. But I suspect you can see that even this tiny effect causes the overall distribution to change, at a very small degree, and that's what you measure in a venturi.
Ok, now let's close this off by looking at your original question.
Take the venturi and saw the top off, and put a flat plate of metal there. I think you can see the overall effect will still be there, but modified. Now you don't get every particle moving "close enough to +X" getting into the pipe, because the ones moving slightly downward hit the plate. So now you get the ones moving along X and the ones moving along +X and a little +Z. So that means there's a little bit of unbalanced force... upwards. That is what you call lift.
Now take the plate off and replace it with the entire atmosphere. Airplane wing.
This may sound all very hand-waving, and it is, but this is precisely how modern aerodynamics is actually calculated. What you do is make a huge 3D matrix in memory, and fill it with a number representing the velocity of that little cell of air. So for instance, in our case we might fill it with 50,0,0, meaning the air is moving 50 m/s to the right (which is easier than moving the plane 50m/s to the left).
Now you say that any cell that touches the object reflects from it. And then in the next iteration that cell bumps into the cell beside it, and so on. And you keep iterating until you (hopefully) get a steady state. Like this:
You can see all of the effects we're talking about in that image. You can see how the air goes down the intake it gets squeezed, and that sort of squeezes it into the restriction. Then it passes the restriction and spreads out again.
This is an example of "do it on the computer" that really works; instead of using mass flow formulas and stuff you just stick the object in the computer and you'll know what really happens.