You can do so with only one touch probe.
If you have a known pin in the spindle (P diameter) (an endmill with flutes won’t work)
And a known touch plate (W wide), and it is aligned with an axis (easy).
Touch the left, note the position, touch right, note the position. The width you will get is W+P. If it’s not, change the calibration in that direction. You can do the same thing in the 90 degree direction.
I think this should be built into CM, it’s fast, it’s easy, and if you have the probe, “should just work” and solve a lot of the problems people have with doing this. It requires -a- known size probe, not necessarily the fancy carbide active one.
To find alignment is more involved, but could be all automatic:
If you probe at the “B” end of A and the “D” end of A, you can now determine
The alignment of the A side to the axis. (because you made a square probe, you also know The B, C, and D alignment) . Probe the D and B sides. You can now adjust the program to align with the block, and properly find the corner of the probe.
Now that you know alignment, you now know the actual distance between A@D and C@D, the difference between that and the length of the side will be the diameter of the end mill (most of the time - you will need to be careful about flute alignment).
This only works as long as the probe isn’t too far out of alignment (probably about 20 degrees or so) - more than that and probing the other side of it will be difficult to hit reliably. You can see if the probe turns a little more that the C@D probe will miss the C side and hit the D side. With a little math you could probe at a guess for the center of the C side, but then it requires a little more effort to find the end mill dimension (but you have all the information you need)