I have experimented with caps over the last 6-7 years relatively regularly, with my HR1s tweeters with caps, bypass caps, and resistors 1st. Then I got into methodically modifying my Torii and CSP3. After about 3.5 years, having more-or-less finished the amps, I started back on my HR1 speakers after updating them to Bob's latest version. I came up with a way to make my crossover plate removable from the outside, and Bob agreed to set it up for me. So now I have been playing with my HR mid-bass driver crossover and internal wires too, while refining the tweeter caps, and doing some final refinements in my amps, mostly resistors and wires.
The difference from your situation is that none of these caps are very high value, and I can use a lot of the same caps on speakers or amps. My mixes on the speakers, and in one important amp position, start with a 3.3 uF caps. Being able to use them on my tweeter, mid driver, and input section power supply of the Torii, I have been able to try a number of caps in multiple positions. And the same with bypass caps...I use them about everywhere I can... so lots of them!
Before getting into the Torii and CSP3, I was playing for years with variations on my speaker tweeters, getting a good feel for different decent caps, several 3.3s mixed with many small value caps for bypasses. On the tweeter I like a 3.8 -3.9 uF value, so I could buy pretty nice base caps, and nicer bypasses without even approaching Duelund costs. Having gotten quite a few nice mid-grade caps for tweeter bypass experiments, I was ready to see how they did when I started bypassing my electrolytic caps in my amps.
I guess you have seen the humble homemade hifi evaluations? I agree more-or-less with much of what he found.
http://www.humblehomemadehifi.com/Cap.htmlFinally, I have had a lot of luck mixing caps in parallel to make a better larger value cap. Some mix better than others, but the cool thing is, with complimentary combinations,
this has always worked for me, the combo cap better than the original would have been in a higher value. I have not tried it, but I suspect even a (good quality) 3.3, and 0.33, and a 0.1 of the same cap wired in parallel would sound quite a bit better than a 3.7 of the same cap...giving more fine detail, space, and associated textures and spacial information. There is something to splitting up the work by taking advantage of refinements from decreasing frequency ranges each smaller cap supports...working on less of the range, so doing it better. In my experience, a decent 0.1 cap will reveal the limited range it works on better than a higher value cap of the same make can. So what some folks call "controversial" or psychoacoustic opinion relative to bypassing, does not match my experience....not even close. For me careful bypassing, no matter where, has been a huge support for ultimate realism and beauty.
Especially in my amps, but also quite useful on my tweeters, I went further, using many 0.047 - 0.0047. Lots of those were 0.01s and 0.022s bypassing 0.1 up to 0.33 bypasses, making notable refinements in complex detail, higher frequencies and space.
Which leads to your question....I think your track makes good sense. Bypassing a lesser high value cap with one, or better in many cases, a sequence of low value caps of a higher quality, can really work.
In fact, to me, when it works, a combo cap with a decent base cap, and better quality bypasses, can be way better in a given position than either of the mix, and I suspect as good or better than many single very high grade caps. I can't say for sure though, as I have not tried Duelund caps, Jupiter Coppers the highest cost caps I have tried.
The great thing about mixing is you can do it to taste. Each good sounding cap having particular good areas of revelation, you can really get refined by carefully mixing ones that excel in some areas, with others that excel in other areas. Trial and error can result in combinations with really nice balance and complexity. And having lots of places to explore bypasses and coupling caps in my amps, it has been interesting how cap's sonic qualities conveyed between amp and speakers pretty well. Not exactly, but pretty well.
Right now for my tweeters I have a Jupiter VT 3.3, bypassed with a 0.22 Miflex copper oil (these are "realer" sounding to me than Jupiter coppers), Jantzen Superior 0.33, and a 0.022 Miflex poly film/copper. Four caps is crazy maybe, but having messed around as much as I have with caps, I have a lot around, and this was a way to get the value I wanted while mixing known cap sounds, and with a nice sequence of values. I got lucky, and it created a beautiful smooth, revealing and complex balance with my ribbon tweeters.
Which reminds me.
In my setup, and with this tweeter, the VTs and Miflex Copper Oils sound more open and revealing, less thick, with the cap ground line going to the positive speaker post...just a heads that testing directions can help.
On my mid driver crossover, after several experiments with other caps, I followed the coil from the hot wire with a Jantzen Superior 3.3, and a Clarity MR 1.5 parallel. The combo is 4.8, pretty close to the original 4.7 cap between the coil and negative post. I wanted a little more liveliness from the mid driver, so Bob suggested putting a low value resistor after the cap. Having tried several types and values, I ended up with a Duelund 10 watt, 3 ohm silver resistor between cap and negative post, reducing the intensity of the cap effect, and opening up the mid driver enough to create a balance with the other drivers that I prefer. I have a few more caps and resistors to try in this position, but am so happy with what I am getting, I have not changed anything for a few months.
All that said, I am ignorant about big film caps that are affordable and mix well. Do you know Jeff and soniccraft? If not, he would likely be a good one to ask for a quality base cap that won't break the bank, and good bypasses. He handles Miflex too!