The JM Donut may not reflect the audio back to the speaker, that is not as much of the point I was making. It was that the path of flow that the HF audio makes through the foam mesh, has many different path lengths. The path length in the center vs the path length to HF that exits at the periphery of the Foam donut are drastically different. The tangent between those two exit points and your ear is what will cause the destructive interference. Now imagine all the other potential paths that the audio can travel and exit through that foam donut at! They are all different path lengths, meaning all of them will shift the length of time relative to one another as it exits the foam. That is the cold hard physics of it. Cannot be changed and 100% exist.
Now, that does not mean that it doesn't do what it says it will do. It also does not mean that it will sound bad doing it. I am only pointing out that it does something that is considered, objectively, to be a bad thing. And if Jay or anyone else tries to tell you it doesn't, they are lying to you or trying to sell you something.
Not having actually read any of Jay's stuff about his donut, I can only see one way to spin it. Correct me if I am wrong. My guess is that his supporting thesis is that the HF energy travels through the foam and exits out of thousands of the different " cells " in the foam. Each one is so close that the destructive interference between any nearby cell is minimal to nonexistent. He is trying to use line array theory for this. The path length between each adjacent exit point is so close that it effectively couples and the time variance from the center to the periphery causes a coherent dispersion of HF audio. That is my best guess as to what he said to support his product. Am I way off base, or close enough?
I would say that at distance, this idea holds up fine. But in the near field, perhaps not so much. When you place a microphone in front of the foam mesh, you now have a very small diaphragm that is only able to pick up a fraction of the cells. Each cell has a different path length and they will propagate over time differently on the microphone's diaphragm. I.E. the microphone's close proximity to a few cells and its ability to hear other cells further away make destructive interference at the mic more possible. This is of course an oversimplification of the potential. The math is where we really learn.
Sound effectively couples when the two sources' acoustic centers are within 1/4 of a wavelength of each other. Beyond that, they start to destructively interfere. A 1khz sine wave has a wavelength of about 1'. So for effective coupling to occur, the two acoustic centers ( the cells in this case ) must be within 3" of each other. No problem in this case if the foam is acoustically transparent to frequencies above 1khz. If not, there will be minimal destructive interference at 1khz, there already is with just the speaker cone alone.
Most guitar speakers are good up to about 5khz though. So what's the math on that look like? A 5khz wavelength is about 2.7" (let's just say 2.5" to keep it easy ). This means any two cells that are beyond about .625" ( 5/8ths of an inch ) apart, will start creating destructive interference. A microphone can hear much more than 5/8ths of an inch worth of a speaker cone. There will be a fair amount of destructive interference the microphone can actually pick up.
Going even higher in frequency, perhaps to the limit of interest for guitarists, a 10khz frequency has a wavelength of about 1", this means that any cell that is beyond 1/4" from one another will create destructive interference. There is no doubt that there is comb filtering going on at this frequency. But why does this not make as much of a difference when you're further away?
The angle of incident is the key. If speaker A is 1' directly to your left and speaker B is 1' directly to your right, as long as you stand still you won't hear anything weird. But as you shift your head from side to side ( changing the relative distance between Speaker A and Speaker B to your head ), you will start to hear the wooshing and weird audible anomalies of that point in space. Now imagine that you stay directly in between the two speakers, but walk away 25'. As long as you keep your head perfectly between them you will hear them as if it was normal, but as you shift your head left to right, you will notice that the change in the sound you hear is different, it will only affect the higher frequencies. This is because your acute angle to the speakers means less drastic of a change in tangent between them. Now as you move further away to say 50'. You may notice that you can't move your head far enough left or right to hear any major change in sound at all. This is because you are now so far away that the tangent between the two speakers when you move your head is minimal or essentially too small to cause a problem.
So going back to what I was saying about placing a mic in front of a speaker and coming to a quick result. If I don't catch that there is some sort of device in use behind the speaker grill, and I place the mic like I normally would, and I don't like what I hear, I need to go back and move the mic. If again I don't catch that there is a device in use at that point, I will move the mic to acquire an expected result. If it's too bright I will move the mic further out into the cone, and if it is too dark I will move it closer to the dust cap. Now you can see what I mean by expected outcomes vs shooting in the dark. If the device does something that creates unexpected results, I am shooting in the dark as to how to fix my problem.
So again, I'm not saying the JM Donut is crap, I'm not saying that it doesn't work, only that it does something, and it may or may not have negative outcomes. If what the mic hears and what you as a player hears are two totally different things, as a live sound Technician ( to keep it real around here ) I have to try and capture what you are hearing. I am not trying to trudge my own ground with regard to your guitar tone. I do use a bit of artistic judgment and flare, but it is usually only to make things work as a whole, obviously, the guitar can't be so boomy and bassy that it walks all over the bass guitar. If your guitar tone is that full and big, I have to do some amount of neutering to make things work.
All that stuff above is what I have rolling through my head when I am placing a mic. Do I pull the mic back off the grill a little to change the angle of incident, do I move the mic left or right, do I move both back and to the left or right? I'm doing real-time math in my head as I place the mic and then come back to move it. The hope is that if I do move it, I made the right assessment and came close enough to solve my problem.
Why can't I spend more time? I have a drummer, to deal with many of the same things for. I have percussion too maybe? Then two guitarists. etc. etc. If I'm a monitor engineer, I have a lot to do in a very short period of time during a festival. You only get one, maybe two opportunities to move any one mic and get a good result. This is why I am not a fan of these devices, because they do not exhibit expected outcomes. They solve one problem and create another; usually.