My makeshift kitchen "laboratory".
When I was doing the inductor and circuit board upgrades on my SDA CRS+'s, I took the opportunity to do further measurements on the 12 uF and 20 uF capacitors used in their crossovers.
For the 20 uF capacitors, my findings concurred with jcandy's that the Sonicap's had the most deviation from nominal value. However, there are other factors which affect a capacitor's performance in an audio circuit besides deviation from rated value. Three of these are transient response, noise performance and dissipation factor. I would expect that anyone doing a serious quantitative study of capacitor performance in stereophonic audio circuits would demonstrate due regard for these parameters. My research indicates that there is a link between these parameters and imaging and spatial properties.
With the 20 uF caps, the Sonicaps had the most deviation from nominal value (20.73 uF), yet sounded the best. The stock electrolytics measured the closest to nominal value (20.13 uF), yet sounded the worst. Sonicap's had the best transient response and noise performance, as measured with an oscilloscope, and the lowest power dissipation factor, as measured with an LCR meter. The stock electrolytics had the worst performance in these three categories.
In this study, the 12 uF and 20 uF capacitors which have been used in three pairs of CRS+'s speakers were evaluated. The 20 uF capacitors were: (1) stock electrolytics, (2) Solen PB series, (3) AudioCap PPMF series, (4) Sonicap Gen I. The 12 uF capacitors were: (1) stock mylar, (2) Clarity Cap SA series, (3) AudioCap PPMF series, (4) Sonicap Gen I.
It is not clear to me why anyone would think that a complete understanding of an electronic component's performance can be gained simply by measuring its rated value. Shouldn't we be primarily concerned with how a signal looks (i.e. maintains integrity) after it passes through a component?
Similar to the results with my SDA inductor replacements, I found a direct correlation was between a capacitor's transient response, noise performance, dissipation factor and its sound quality relative to other capacitors. The AudioCap PPMF's were close in measurement and in sound quality to the Sonicap Gen I's, but the Sonicap's provided better spatial rendering, image weight, overall detail and bass performance.
As I noted in my SDA SRS 1.2TL Sonicap upgrade thread (here), Soniccraft's owner, Jeff Glowaki, tried to discourage me from replacing the AudioCap's with Sonicaps. He specifically said that the Sonicap replacement wouldn't be "worth the money" and he specifically recommended bypassing the AudioCaps with small value Sonicaps. However, I politely declined his more cost effective recommendations in favor of consideration for the favorable experience reports of other speaker owners who had made the switch.
I would add the caveat that higher quality capacitors with higher measured performance may not result in an audible difference or improvement in some audio equipment. This is particularly true with high quality capacitors constructed of the same dielectric material.