It doesn't say one at the soffit. It says "the upper one being located at the roof line". It's a good thing it's in the appendix and not a rule?
I guess the phrase "An assembly considered to be installed in an acceptable manner is one which..." doesn't mean other assemblies are not considered to be installed in an acceptable manner. For example, it appears from your comment that you consider the upper support at the soffit to be acceptable, even though that is well below the roof line. Maybe two supports are acceptable when one passes through a structural member.
OK we may be splitting hairs here but the depth between the top on the roof and under the soffit is max 6" so my at the soffit comment is at the roof line in the photo. Now if the drop from the roof line to the soffit was a foot or 2 than I would expect the first mast clamp above the soffit at the roof line as you correctly point out. I get your point and just the text of my reply misses that. I suppose if the AHJ accepts that other assemblies meet the necessary assembly strength then great but we take a more literal approach and 2 clamps is 1 shy of compliant.
My challenge to the AHJ is by what criteria do they accept 2 clamps? What engineer or alternate method do they accept less than 3 clamps? How did they arrive at that conclusion and under what risk model does it work? What mast manufacturer says 2 clamps meets their design parameters? You see I don't disagree that your point is valid I just would ask how another AHJ arrives at a lesser than appendix B install? How did they come to accept 2 mast clamps instead of the 3 described in the CEC.
I also expect you know that appendix B is just an explanation of the rules and not code.