Oh right, I for got all about the update for the bypass testing.
All of these results are relative to one another, starting with the cheapest.
JB JPX: Adds additional detail/clarity, creating a more focused/precise soundstage, without changing the overall tone.
Gen II Sonicap: Surprisingly similar to the JB. Making the JBs an excellent alternative at a similar price, considering these are not longer available.
Gen 3 Sonicap: Compared to the JB & Gen II, it's still quite similar overall, though just a little less detailed, but offers a wider soundstage than either the JB or Gen II cap.
Duelund 0.1uF Copper: Better detail, much like the JB & Gen II, but with a much more open soundstage (both width and depth) than either previous cap, while still adding that additional richness/warmth from the copper. It really needs 400-500 hours of burn-in before the soundstage depth, finally develops.
Duelund 0.01uF Silver: Very similar to the Duelund copper, but more detail that can be a little edgy or grainy sounding by comparison, if not a bit "artificial", IMO. (though likely dependent on the main cap it's bypassing)
Soundstage is a little narrower than the copper in my system, but is identical in regards to depth.
With the JB and Gen 3 Sonicaps, they're very similar overall, and are mostly playing with differences at the edges, and really depends on where you want the improvements.
The Duelund' caps are impressive, especially once you get past the roughly 3-4 weeks of non-stop burn-in...
My preference it definitely for the Copper over the Silver, as the copper bypass caps just sounds more rich and natural, but if paired with a "dark" sounding cap, the Silver may be the better option in such a case.
I'd eventually like to compare the Duelund "CAST PIO" copper as well as the tinned-copper caps, and maybe throw in the Miflex KPCU and KFPM caps also, but that will likely be for sometime in the future when funds/time allows.
Danny did the same tests with his NX-Tremes in his system a few weeks ago, bypassing the main Miflex cap in the tweeter circuit and came to the pretty much same conclusions as I did.