I've used
these whenever I've done bypassing in the past. I've only ever bypassed the power inputs of opamps before to avoid osicllation as well as a directed bypassing of a main filter cap in a 39V regulated Ampex power supply.
There are other things at play in bypassing which has led me away from making it a blanket practice. GCalo and I have been dialoging about this offline and I don't have the same background experience as he has so its not as comfortable for me because I get hung up in issues like the biasing (not tape machine bias but signal bias in the power path), ESR and such...there are just other things at play. There's a little discussion on it in
this thread where I was working on the mods and upgrades for my Soundtracs MX power supply. I suggest you look at it because it was an eye opener for me as to a bigger world of things at play that highlight the mass interdependence of components in a circuit and how each component doesn't live on an island, and also how, depending on the architecture and type of system, certain mods or specs or issues just may not be as relevant..."say man, that blender in your Chevy is really...something..."
MY bottom line is that I've gotten more into the mindset that when it comes to modding I just simply don't have the understanding and knowledge to
really assess what is a worthwhile or necessary mod NOR the ability to really test the mods to see if they were effective or possibly even detrimental. Its all WAY over my head. So I that's why I tend to stick to replacing eletrolytic caps with like values. I guess I'm conservative. There are exceptions of course, and I'm comfortable enough where I can identify a cap that would benefit from an increase in voltage (for instance a 50V cap in the path of a 48V phantom supply is too close for my comfort), and yes it doesn't
seem to be a bad idea to increase the value of supply filter caps but I'm careful to ask so I can confirm which ones are the filter caps and so on. When recapping 20, 25, 30+ year old gear you can generally assume that the caps you put in will have better bandwidth performance, and being able to replace with higher temp caps is a nice insurance policy too. So there are potential sum gains in just replacing like values with like values. Where I have incresed the capacitance I have done so on specific instruction from somebody I really trust, and it always seems to boil down a pattern for me that there really aren't any blanket givens in upgrading electronics, except for going with higher temp caps than 85C when able, especially in power supplies.
If there are other blankets then I'd like to know what those are.
So that's my 2p, and take it with the grain of salt that I'm conservative on this...some of that comes from people I respect telling me "why are you messing with it?" I've gone through a phase where I just gutted stuff...I remember the point when I'd pulled all these caps off of a Tascam PCB and just for poops and giggles checked to see how far they'd drifted in value...every one of them...every single one of them was a 20% cap and was within 10% of spec...that was out of a 25 year old unit. Now the ESR may have been way off...that I don't know because I don't have an ESR meter and I'm pretty sure the Nichicon KT series caps I put in were cleaner, had lower ESR and better bandwidth, and I knew after replacing them I didn't have to worry about the caps going bad in the next years, but it certainly made me stop for a second and think "what am I
doing?"
Power supplies are a different story. I just think its a good idea to recap those after 20, 25+ years as an insurance policy and you never know what might be helped. That supply literally feeds everything in the unit...garbage in, garbage out and unless you have a way to analyze the ripple in the rails then why not replace the caps? Most supplies only have a few caps in them anyway...even larger multirail supplies may have just a dozen or so. Recapping is a good way to get your feet wet because its not a big project, there may be immediate gains, many of the caps are bigger and just easier to handle first time out rather than being these little pellets, and many times the supply PCB is easier to get to than some other things.
Bad caps in the audio path tend to result in crackles and pops. Bad resistors tend to result in more hiss and "thermal noise", but when I recapped the supply for
my Tascam M-520 there was an
immediate and very noticeable drop in the noise floor, the "hiss", in the mixer. Maybe one of the filter caps was going bad and that was driving the rectifiers nuts...I don't know.
And not all supplies are created equal. I thought for sure that there'd be ripple issues in the 40 year old supplies in my Ampex MM-1000...I have a 50MHz analog Tek scope and I couldn't see anything except environmental "noise" when zoomed all the way in...very clean...stock and 40 years old. Will I recap them anyway? Sure but its not a huge priority at the moment.
Now, that's all I'm going to vomit at the moment and as I try to emphasize here and there I'm a hobbyist...I have no formal education pertinent to electronics...the worst thing that could happen with the above is somebody takes it and goes "well this guy said blahblahblah so that's the TRUTH!!! 
Take what works and throw out the rest. These are the epitome of opinions I've expressed here. Do your own research. Ask questions. Learn. Experience. Find good people to mentor you. Have
fun.
I considered getting the Nichon "audio" capacitors but the values are limited. Anyway, I'm at least going to get the PSU caps and do that for starters. May just have to take my time on this and research.
Steve, which series of caps were you looking at? I use the KT series. They're light blue in color,
105C caps that are "Audio Grade" for whatever
that's worth. I just like the idea that they were designed (or are at least
marketed heheh) with audio in mind and they are hi temp, come in lots of values and are affordable. A good balance/blend for me and Mouser carries them. They aren't the "MUSE" caps that people rave about, but those are 85C caps anyway...I DO use the MUSE caps for bipolar audio path stuff...those are the ES series and are indeed more limited in the values available. I tend to use the PW series for non-audio path stuff and in the power supplies...VZ series if I can't find the right value in a PW...bigger stuff or Axial stuff I'll tend to use Cornell-Dublier or Vishay caps...really big stuff I like the Mallory computer grade caps...
If the electros are signal-path, you will get a larger benefit by replacing them with nonpolar (bipolar) electros.
Oh Msh I
wish I
wish I
wish I could sit down with you with a schematic from a typical Tascam mixer strip and circle the polar caps that could be replaced with non-polar. I think I understand what you are saying but I always thought polar had to be replaced with polar, non-polar witrh non-polar; that the reason a certain type was selected was because that's what was needed...why wouldn't "they" put a non-polar cap in instead of a polar cap at the design stage if there would be a benefit? It seems too easy in a way...
Thanks for the pic GCalo. I see. So you do them all with .01 mFd caps? Even the power supply? 
