I think I finally understand what yaw is about. I must admit having been confused about it in the past and even giving some misleading, incomplete advice. I will try to explain what I now think and see if anyone thinks I'm right.
Yaw, pitch and roll refer to terms used to describe the change of attitude of a rigid body such as a ship or an airplane, and in view camera photography, we use tilt, swing, and yaw to describe what happens to a standard. (Identifying which corresponds to which depends on how you set up the reference system.)
As we use the term yaw, it refers to rotation of the standard about an axis perpendicular to it, e.g., the lens axis for the front standard. Yaw may result from performing a tilt followed by a swing, or, vice versa. Exactly what happens depends on on how the axes for tilt an swing are attached to the camera frame.
Yaw could clearly be a problem for the rear standard because it would change the orientation of the film frame with respect to the horizontal and vertical. But I never saw why yaw would be a problem for the front standard. After all, the image is symmetrical about the lens axis, so you you can rotate the front standard about that axis to your heart's content without seeing any difference.
What I've finally figured out is the following. It is not yaw itself that is the problem for the front standard. It is what caused the yaw in the first place. Let me elaborate.
Suppose you apply a tilt. If you have axial tilt, this will be a rotation about a horizontal axis through the lens. But even if you have base tilt, you can adjust the position of the standard by rise/fall or movement along the rail to accomplish the same thing, rotation about a horizontal axis through the lens center. So let's assume that is the case. The question then is what happened to the swing axis. For some cameras it may remain vertical where it was. That will generally be the case for cameras with axial tilt, but it is not generally the case for cameras with base tilt. For the latter, the swing axis stays fixed in the standard and hence is tilted with it. This is usually phrased by saying that in the good case the "point of attachment" for tilts is "above" that for swings, but I've never understood exactly what that is supposed to mean, so I just try a tilt and see where the swing axis goes to be sure.
What happens next is a bit subtle. You do get yaw in the bad case where the swing axis has moved, but something else happens that is more complicated. Consider what happens to the hinge line. You hope you can move it through tilt followed by swing so it coincides with where it should be in the desired subject plane. After the tilt, the hinge line is horizontal. In the good case, as you swing, the hinge line rotates in the vertical plane parallel to the image plane about a fixed point---that is harder to see than you might think, but it is right. If you got the tilt right, which turns out to be feasible, you can rotate the hinge line to where it should be by a swing. The amount it rotates is related to both the tilt angle and the swing angle by a complicated formula, but the important thing is that for fixed tilt angle, the swing angle and the hinge line rotation angle determine one another, and you don't need to know what that formula is.
In the bad case, whatever tilt angle you choose, the hinge line will move in its entirety, no point in it remaining fixed. I don't see in that case how you can possibly predict in advance just how much to tilt in the first place, without doing elaborate calculations. You would have had to have overshot or undershot the tilt by exactly the right amount, and thaat would require knowing the formulas. So you end up having to go back, adjust the tilt, readjust the swing, etc. Of course, since in practice, you never get anything exactly right, you may have to do that anyway. But if you go about it right, fewer adjustments should be necessary.
I'm sure someone will say that the effect I'm talking about is an obvious effect of yaw, but it wasn't at all obvious to me. I had to look exactly at what happened before I saw the importance of the hinge line for this, and the geometry was not that simple. It only occurred to me when I was thinking about something else entirely. It seems more enlightening to me to say that both yaw and the effect on the hinge line result from the same basic cause. In any case, I think it is well worth understanding the importance of the hinge line in all of this. Of course, if you already have a yaw free camera, you will know from experience that something is better, in which case you may not care why, but if you don't have such a camera, which was true for me, it may take a while to figure it out. Anyway, I think I now understand the reason why many experienced view camera users extol the advantages of being yaw free.
Let me mention as an aside that usually base tilt cameras will work just fine in this regard if you just do the swing first. At least for all I've looked at, including my Toho FC-45X, the tilt axis doesn't change when you swing, so the good case becomes the bad case, and vice versa. I didn't know this because no one seemed to mention it anywhere, so it wasn't until now that I realized that I should swing first and then tilt. For the axial tilt cameras I've looked at, you run into the same problem if you reverse the roles of tilt and swing. I haven't looked closely at any cameras purported to be yaw free, but for that to be really true, the swing and tilt axes would have to be entirely independent of one another. I'm sure that can be managed by an appropriate mechanism, but as far as I can see few view camera manufacturers bother.
Any comments would be appreciated.