I would not specifically call them "bass heavy". The way their response below 1kHz is shaped, when I measure them on my head with various in-ear mics techniques, or when they're measured on various test rigs, is quite close to Harman / K371 for example.
Now someone's subjective assessment of a pair of HPs' bass response has a lot to do IMO about what happens
above 1kHz, as it's the overall spectrum that matters. A lot of dummy head measurements tend to show that the APM's response is quite conservative vs. various target curves in the 1 to 4-5kHz region.
This alone could make the response below 1kHz stand out more. But as I said,
and interestingly I'm not the only one, using in-ear mics show that the response in that region isn't quite as deficient overall as what test rigs would suggest (albeit still quite conservative, and I personally find them deficient above 3Khz in particular - but not much below) :
https://www.audiosciencereview.com/...k-headphones-200-300.16005/page-3#post-828653,
https://www.reddit.com/r/headphones/comments/os73cl/airpods_max_impressions/
Above a few kHz the APM's measurements are all over the place. Difficult to extract any valid information here. Personally I'm getting too much trebles response past 6kHz or so overall.
In general above 1kHz it's entirely possible that the APM exhibits quite significant person to person variation if we were to measure their FR at the listener's eardrum.
What I feel however, is that two decisions that Apple stuck to during development, ie the headband to cup attachment pivot mechanism (and more specifically the fact that the headband joins the cups at the top and requires the latter to be spring loaded to compensate), and the fact that the earcups are symmetrical front to back (as Apple originally floated the idea of making the APM L/R reversible and stuck to that design even when they didn't pursue this idea further), made them a lot harder to effectively tune and develop past 1kHz than they should have been. My suspicion here (and it's just an idea that I'm floating) is that these two elements combined may make them couple with test rigs, which aren't that anatomically correct
around the pinna, in a way that isn't a superb match for the average human (notwithstanding another idea that I'd float : I suspect that these design decisions compromise quite severely the person to person FR consistency of the APM).
Combined with Apple's usual secrecy and possibly the lack of mass-testing with real ear measurements
and the final deliverable product (I have very little doubt that Apple spent a considerable amount of time with proxies to determine on their own what listeners preferred the most) that they could have done otherwise I'm not too surprised that the tuning above 1kHz may sound a little bit off to quite a few people, may show quite a bit of variation across listeners, and may result in the APM's sounding "bass heavy" to
some people.
I still find them much better tuned than most other over-ears ANC headphones I've tried.