Parallel Capacitors-Theoretical Question

Since the engineers at Advent used a single 13uF cap, my guess is that they didn't want the high order harmonics. Using 6.2uF & 6.8uF would alter the harmonic structure somewhat; and using 12uF & 1uF would alter it more. I guess the choice would depend on what kind of tweeter is in the Advent, e.g., silk dome or metal.

Jburidan-very good insight. I think the Original and New Large Advent tweeters were more silk than metal and were described as the "fried egg" tweeters. They had their own unique sound which I think defined the character of the speaker. The crossover was a simple 6db first order type. Advent later changed the tweeter to a black poly unit and used a different crossover (3rd order), which in my opinion ruined everything.
Can anyone explain how adding a bypass capacitor would affect the sound? Is that what the old "Vitamin Q" was about? Why was it described as a mixed bag earlier in this thread?

That 13uf electrolytic cap is 20% tolerance, a 12uf would probably be fine.

I gather that the results of adding bypass caps are unpredictable. Some reviewers (Auw Jimmy, Jon L. & Tony Gee) make the analogy of usng spices in cooking. They experiment until they hit upon a sound flavor they like.

I was really trying to get away from experimentation, it gets too costly. I realized that after rolling op-amps into my phono preamp recently. Before I knew it, I had spent nearly $300.
Capacitors are a bit more complex I think. There are more issues than their face value that can affect them. For example, Mundorf interleaves their capacitor's windings to cancel inductance. After I read that, I wasn't sure what might happen with 2 caps in parallel. Would they induce each other and do strange things at a particular frequency? There's also ESR to consider.
Although I realize many designers make their component choices from an economic standpoint, I still respect those choices as being the best available at the time to get the job done. After all, isn't engineering the art of compromise? That's why I wanted to keep the 13uf value, or get as close to it without causing other problems. It would therefore seem that a good 12 uf capacitor would be the place to start, maybe add 1uf to see how it changes things, and maybe add a bypass 0.1mf after that.
Thanks for your help guys. If anybody's interested in the final result, contact my email and leave me a message.