roskodan
Headphoneus Supremus
I'm swimming upstream on this, I know, but I take a different lesson from this: don't attempt to mod headphones unless you're prepared to ruin them.
Headphones, by and large, are a completely manufactured, often machine-fabricated product. Either they're too cheaply made to take apart or they're too expensive to risk it. A headphone company is in the market to compete. It's ridiculous to assume that a headphone mfr is not trying to make the best sounding headphone at the lowest price. So, they've already optimized the audio quality. It may not be to your own particular preference, but they tuned it to sell, period. Likewise for the manufacturing - pricing is so competitive worldwide, that every effort is going to be made to reduce manufacturing and fabrication costs. That doesn't always mean the result is going to be easy to dis-assemble and re-assemble. More often, it won't be - repeatable connectors/fasteners are always more expensive.
I've been a bit dismayed at how this forum section has seemed to transition from electronics and electrical DIY to modding headphones. It seems to be the tide that's sweeping around here lately. There have been some notable successes in modding with the Fostex planars, but those are a glaring exception, not the rule. This thread is an example of that.
I'm sort of hoping that the tide will go out on this headphone modding wave and we can get back to actually creating entire products with DIY instead of the latest blu-tack and sock mods. JMHO, but I'm probably in the minority right now.
I understand your view, but a $2,000 headphone should not have these inconsistencies with the frequency response in the treble and neither should it have those issues with the pads.
in fact even without measuring there have always been customers complaining about the sound deviations, from sample to sample, across the lcd product range
1) i suppose the OP thought his measurements applied to all the lcd3 manufactured, hence the need to do the mod as the only possible way in getting a '$2000' sound/measurement/quality (or it was purely a curiosity/scientific thing, perhaps out of warranty as well?)
2) while the truth is that each lcd3 (or any model in the lcd series) is different in sound and measurement from sample to sample (even the hd800 do, each have its own graph which is visibly different, although the margin of 'error' is much lower), as they are manufactured in a niche, home made, kind of process, in other terms, with loose or no strict standards, low tech manufacturing with higher margin of 'error' (this is not a bad thing, it's just how niche works)
3) that's why there is a warranty and/or return period, so if the product does not match a '$2000' standard, which is also a highly subjective matter in audio (not to mention gear matching, and the effect of the whole chain on the headphone's performance/sound), you can have the manufacturer to deal with that risk, no need to do their job
other than that, for diy-ers, it's very sound to be well documented and informed to limit unexpected risks, like the disposable diaphragm assembly, while diy-ing

also i doubt a lcd3 with better measurements would be proportionally better sounding, or even make a meaningful audible difference, that's just how the lcd3 treble is, it takes another headphones to change this i suspect, since my old lcd3 2013 had that kind of hot treble compared to my lcd2 and my lcd-x
this thread reminded me of some folks that do not like a piece of gear and then start buying oils and dusts, so to speak, instead to just get rid of the d***ed m*f*, LOL