What works best for me is to have a reference set like a stock set of T50RP's, a favorite modded set, or LCD2's for comparison with my current modding analysis. I choose 3 or 4 songs of different genres that I know very well. I sound level match my reference and test cans with my SPL adapter for each song so I know where to set the volume knob on my Dacmini during comparisons. I listen to certain sections that I am trying to tweak like a bass line, vocals, snares, or high hats and play the same segment for 30 to 90 seconds - and alternate from one set of cans to the other until I can discern the similarity or difference I am trying to tweak. I repeat the process for each song and make notes when needed. Then, I make one mod change and repeat the process...over and over and over until satisfied or frustrated. I can open, mod, and close in 5 minutes, or less, depending on the difficulty level of the mod. Since I leave the plasticine and Paxmate mods constant, each mod is usually completed quickly.
For example, today I compared 3M Nexcare flexible clear tape X layers X grid placement X reflex dot against my LCD2's. Three hours later I have achieved the best I think I can accomplish with standard materials. The LCD2's remain superior in every way from width of soundstage to airyness to clarity to bass quantity an quality. I think I am at 85% of the LCD2's with my current mods to the T50RP's. Considering they are semi-closed and the drivers are 1/3 the size of the LCD2's, I think that's a pretty good outcome. In fact, if I compare stock T50RP's, the modded set sounds better and I can tell them apart when my wife does the old switcheroo test with me as the test subject.
Here's what I've got:
Plasticine and Paxmate in the usual places
Stock black bass port felt in place
Stock ear pads with 1 donut of Paxmate underneath
White driver felt removed
2 layers of 3M Nexcare flexible clear tape on entire back side of drivers
Individual squares of this tape at 2, 4, 6, and 8 grid spaces
Large reflex dot on center grid space
I tried every combination of tape layering and placement from 1 to 4; and 2 layers with individual grid spaces added in a cross, 2-4-6-8, 4 corners with and without the center space covered. I think there is a curvilinear relationship with my setup and my ears: 2 layers is not enough; 3 to 4 layers is too much; and 2 layers with the 2,4,6,8 + center reflex dot = "The Goldilocks Zone."
I think I have exhausted the multiple permutations of these mods and extracted all the goodness I can; unless some of you good guys toss a banana or two! I'm open and considering some different directions...
I think that's an important observation. My hearing also rolls off around 14khz and this could be muddying my perceptions. I'm also wondering if I'm being to lackadaisical in my auditioning technique. What I most commonly do is listen with mod A for an hour or so. Then I apply mod B and listen for another period of time and make my judgment after that. But I'm inconsistent in what I listen to and, because I've been listening to a lot of 60s and 70s R&B (Syl Johnson, Sly Stone, etc) the mastering is all over the place. So my baseline isn't particularly reliable. That's compounded by the fact that my SPL is initially quite different between listens, until I adjust my pots, which is throwing in some more bias.
I'm going to have to listen longer and more intently. By I'll do it, if it's for the betterment of science :)