@shadorne You say "problems’ as a speaker designer I say "possibilities."
The directivity of an AMT is different than a dome or ring, but if you are going to attack it, you will attack all line or planar sources, including ESL speakers from Martin Logan or Sanders, Quad, blah blah. not to mention most horns as well.
You are conflating compression with directivity and lobing. Three different things.
Let’s get compression out of the way first. When you add X dB to the input but the output is not uniformly X dB louder. About the only reviews that have this routinely are the one’s from
speakermeasurements.com
So I stand by my original statement, the best AMT's can compete with the best of any other type of tweeters in terms of compression.
Let’s discuss the rest. Lobing isn’t really caused by a single driver but how it interacts with another. It is caused by one driver interfering with another at different angles over the frequencies at which they are both working. This can be pretty complicated as the acoustic distances between 2 drivers varies in three-dimensional space. This is why lobing is a 3D problem, and why many manufacturers align their drivers in a vertical line, to minimize the complex and unpredictable (but not impossible to simulate) nature.
A single driver can not lobe. Single driver and single panel ESL speakers are immune from this behavior but any 2 drivers can lobe, it is not an issue of planar vs. dome at all. It’s an issue of the acoustic centers not being aligned and crossover design choices.
Any time you have 2 or more drivers which overlap in the frequency domain and are not coincident in 3D space lobing can occur. With a theoretically perfect crossover, lobing disappears, but no one has made this yet. Admittedly the Joseph Audio 100 dB/octave crossovers come close. Co-axials are another nearly perfect solution to this problem.
Lobing is sometimes deliberately caused. For instance, the super-expensive, super-ugly B&O Beolab 90 takes advantage of this, and uses DSP to stagger the delays between similar drivers. Also a technique used in professional arrays. In the Beolab only dome’s and cones are used, but it’s been demonstrated many times they can "lobe" very well.
So with that out of the way, neither very broad nor very narrow drivers are "best."
Narrowing the angle of radiation can be VERY beneficial. Time domain problems, and therefore frequency response, at the listening location are greatly reduced by using larger diaphragms and have the perception of transparency and neutrality in spades. Anyone who has spent time listening to a good AMT or larger ribbon driver will attest to this.
Sure, if you are designing a speaker that needs extremely broad radiation in the vertical and horizontal plane, a broad frequency, planar driver is not going to be for you. That’s a reasonable trade-off, but neither should be judged by the Audiophile Gods of All That is Good as an absolute measure of "better."
Unfortunately, I find your prejudices unjustified. You should of course buy and live with whatever drivers you want to but AMTs in general deserve a better treatment than you are giving them.
For the top end, I really encourage anyone who can listen to the Gryphons to spend some time with them and compare to some of the low-end "giant killers" being touted as superb.
Best,
E