Why does my 2hp kick module lose low end when I turn decay up?

Nine times out of ten when a kick does this it's because you've got two time constants fighting each other. The pitch sweep and the amplitude envelope are on seperate caps, right? If the sweep cap discharges way faster than the main decay cap, then on short settings they overlap fine and it sounds tight. Crank the decay up and the sweep bottoms out early, so the resonator settles at some fixed pitch while the amp tail keeps ringing. That fixed pitch ends up high and thin, and the click transient stays audible because the body never develops underneath it. Papery tail is exactly what that sounds like. Easy way to confirm without a scope: patch the output into something with a tuner or just use your ears. If the long decay version starts at the right low note but slides back up or sits at a constant higher pitch through the tail, that's your sweep dying before the envelope. If the pitch stays put the whole way down but the volume curve sounds wrong, then it's more likely the decay cap is loading the tuning node and dragging the voltage around, which is your other suspicion. If it's the sweep, the fix is usally one of two things. Either scale the sweep cap so its tail stretches with the decay pot position, or hang a buffer on the tuning voltage so the decay network can't pull it. I had almost this exact thing on a 909-ish build and a single op-amp follower between the tuning node and the sweep discharge path sorted it out completly. Also check wether your click or accent stage is gated by the same envelope or runs on its own short one. If it's free-running, a long decay leaves it sitting out in the open with no body under it, and that alone will make the whole thing read as clicky.

I'm not buying the two-caps theory here. If the sweep and decay were fighting each other, you'd hear some weirdness across the whole knob range, not a clean cutoff at one spot. A hard threshold past three quarters is classic pot behavior in my experiance. I had a filter module do almost exactly this, fine til the top of the travel then everything went thin and fizzy, and it was the wiper losing contact where the track wears out, not the circuit design. Work the knob back and forth a bunch of times and see if the point where it breaks moves around. If it does, spray it and call it a day before you go adding buffers to anything.

Had somethig similar on my own kick clone and it turned out to be way dumber than expected. The long-decay position on my pot wired up so the last bit of travel basially tied the wiper to the end lug, which dumped extra resistance into the timing network and the whole thing went thin right at the top. Desoldered the pot, checked it on the meter, range was fine, but my wiring had it backwards for the taper. Flipped the lugs and the long tail came back. Might be worth pulling the pot and measuring across the travel before you chase the envelope around the board, takes five minutes and rules out a whole pile of stuff.

On my build the same symptom turned out to be the decay voltage leaking into the resonator bias instead of just the timing cap, so the longer the envelope stretched the lower the loop gain sat, and the fundamental stopped ringing entirely while the attack click carried on untouched. Worth tracing where that pot wiper actually goes on your schematic, because if it feeds anything besides the cap, a long setting changes the tone and not just the length. A meter on the tuning node during a slow trigger would settle it in a minute.

Update: checked the pot travel like the wiring issue poster suggested and there was actually a dead spot right near the end of the range, so I swapped in a different taper and the dropoff is way smoother now. Still getting thin bottoms on the longest decays though. Would measuring the decay voltage at the resonator side be the next step, or shoud I just swap that timing cap first?