Depends. Do you want lowest ESR? That can be a problem with some supplies. Lowest DF? Ditto. Smallest package size? That can be useful in some designs, but the tradeoffs may work against you in others.
The datasheets are on Wima's site and bear looking at. That way, you can decide what represents the best tradeoff for your design goals.
For DC blocking in the signal path, I like the MKP10. I used them on several of my no oversampling Philips CDPs and preferred them over some far more expensive models (different Audyn Caps, Mundorf MCaps etc.).
The only problem with them is their size – they are huge! And it can be quite difficult to buy them in larger values.
An interesting comparison can be found on the following website:
Do you know where the caps are used by the author for listening evaluation. I thing placing the cap in rails decoupling or DC blocking in signal path may lead to different result. DC blocking in signal path may have larger impact to sonic performance?
I had MKP 10 over 15.000uF LCR PSU caps in my old lateral Mosfet JLH 100W amp. When I made a restoration, I pulled them out and the amp became subjectively more coherent in the treble, by a good margin. Andrew is right, we must know we don't upset an amp that has been designed with lossy filter caps when making additions.
I think it is the memory effect (an unspecified parameter) of the dielectrum that affects the sound quality of DC blocking coupling capacitors. I would like to see a method for measuring this effect, perhaps using some bridge circuit.