Was waiting to post until I had a chance to do an accurate A/B test...
About a week ago, I swapped out a capacitor in a friends speaker.
Speaker is an SB Acoustics coax in a DIY enclosure (ported - approx 13 Liter). The crossover is wired as per SB Acoustics recommendation. Quite a lovely sounding speaker.
My friend ended up with a pair of 15uf 600V M-Cap capacitors. We wanted to see what, if any difference, there was between the upgraded capacitor and the more basic 15uf 250V Solen capacitor which were used throughout the build.
The capacitor which was swapped out, was the 2nd of two in the path to the tweeter.
Went from a "standard" audio grade cap to a high end cap. (Solen retails for about $5.00, the M-Cap about $90.00 - considerable cost difference) It should be noted that the driver itself retails for about $60.00...
Only one speaker was done, the other was left as it was originally built. We did this so next time we met up, we could A/B them in a bind test (for me...)
Listened to several amplifiers and recordings (run through preamp in mono)
One speaker had a bloated, boomy bass, mids were set back and the highs got grainy and shrill. The other, while not digging as deep in the bass, was more controlled. The mids were far more prominent and the highs were smooth, clear and details easily identifiable. What surprised us both was how the bottom end sounded so different. Don't understand how that could be happening, but it was definitely there! And I should note, after the second speaker was updated to be like the first. They sounded exactly the same.
Now, the speaker with the original crossover was by no means a bad sounding speaker. While these sonic differences were easily heard, this was about the relative differences in the speakers. The speaker with the upgraded capacitor had better control across the entire frequency band. This was a consistent finding with every amp, and this difference got smaller as the amps got better.
Now, some might prefer the sound from the original, possibly finding the other too polite. Others may not hear enough of a difference to care.
Has this convinced me that I need to replace every capacitor with high end audiophile grade capacitors? No. None of the drivers I "play with" warrant this level of component, even though there was an audible improvement. Now, if I was going waaaaay up the food chain with drivers, and I had completed a design I was absolutely thrilled with, different story. 5K invested in a well designed, built and executed DIY speaker will perform at a level reserved for much higher costing speakers.
All that to say, there was a difference. Is it worth it? Well, that's up to the listener...