Start by pointing them straight ahead with no toe-in. Toe them in incrementally from there. I have not worked out the degrees of toe-in for my speakers, but they are only toed in about 1/2" from corner to corner vs. the front wall. That gives me the best imaging and frequency response in my room.
What is to disagree with? Amir tests samples of products that are out in the wild and posts objective results. Manufacturers can address those results, explain why they value subjectivity over objectivity, remain quiet, or go ballistic. (I don't recommend that last one.)
Manufacturers such as...
One of the cool things about the NFS is you can load its data in Excel and recalculate against a different reference axis (thousands of individual sweeps are included) and get a good idea of how on axis FR would look. And the M-Lore looks... worse at the woofer axis than what Amir published...
That is part of it. The additions to the port will show as a small deviation in FR, probably in the midrange.
Another dimension is impedance. A leaky box will definitely change the impedance curve. Whether it matters or not depends upon how it is changed.
A third dimension is port, or in this...
As a former DIYer who has... err... made mistakes in the past, I can tell you it will mess with the impedance plot. You may not notice if you do not have a good simulation of how it should look. Of course, modern software tools like VCAD make this easy. This is likely the largest area of...
@amirm , I wonder if the 1.2K dip is the shadow flare issue? My R3s showed a similar dip centered there, and after I pushed in the shadow flare rings all the way, the dip disappeared completely.