Experience and try for yourself. This is the only way to know for you .@grannyring A lot of people are of this opinion. I've just found that the more transparent the system is, the more you hear that paralleled caps don't have the desired effect. But I'm sure neither of us want to argue about it- in fact I suspect my viewpoint on this topic could easily be unpopular.
People tend to think that since we make some pretty esoteric products like our fully balanced triode zero feedback OTLs, that we might be doing more listening than measurement or something like that. But to get to where we are takes engineering- and that requires math. If I can't make it work on paper, I have trouble with the idea that it will work in practice. I don't like it when things work and I don't have a good explanation- I remember the first time I heard a power cord make a difference and it really bothered me until I sorted out why. I'd love to be proven wrong on this capacitor thing but so far that hasn't worked out- we certainly *have* tried it and more than once with exactly the same results, 22 years apart. It doesn't work out on paper either, so I've maintained this position for a very long time.
One area that might be a variable is output transformers. They have capacitive effects of their own (both inter-winding and to ground) so it may be that in a transformer-coupled amp there is a benefit as one is compensating for a transformer. I've not worked that out as I try to stay away from transformers if I can.