Advice Needed On Recapping Power Amp

Going to higher capacitance means also higher inductance hence slower response to rapid current demand.   It will also reduce ripple and that might not be a good thing, since very low ripple, especially when combined with low ESR means very high and very narrow current spikes charging capacitors and that might couple to any inductance in the rest of the circuit.  I would agree with  kosst_amojan and would not try on my own to redesign an amp.  In some cases it might work in other it could make it worse (or damage rectifier).

Bypassing slow large inductive caps with fast low ESR cap might be not the best idea, since it creates parallel resonance circuit that might ring.  Better option would be to get low inductance electrolytic caps (like slit foil type), but they are expensive.  Better yet is not to "redesign" something that works just fine.  There is also an issue of reselling such "improved" amp.

15 year old capacitors, unless they are in very warm place, should be good for another decade.

Timlub, Things that look simple not always are.  There is no such thing as "Linear Power Supply", unless Linear means "Unregulated".   In reality it is primitive unregulated switcher that operates at 120Hz and switches at max voltage.  Current from mains is delivered to capacitors in very narrow current spikes of very high amplitude. Amplitude of these current spikes depends on transformer, power cable, ESR of capacitor etc. while width of the spikes depends on the ripple amount that, in turn, depends on the load current (definition of the switcher).  In addition to current spikes charging capacitors you get also very narrow spikes at the peak of the (rectified sine wave) voltage when rectifier starts conducting for a moment in opposite direction to quickly snap back.  To widen this spike there are "soft" diodes (like HexFET), that conduct fast but snap back slow.  I'm trying to show that operation of such primitive supply is not as simple as it seems and the fact that larger cap works means only that safety margin for rectifiers still exists but was reduced and might be very small.  If we believe that designer put unnecessary margin by mistake that's OK, but I would ask him  :)

As for inductance - electrolytic caps are inductive.  Inductance of such cap is proportional to capacitance.  Increase in capacitance means increase in inductance.   Believe it or not, but power supply caps are in series with your speaker (circuit is closing thru them) and their inductance can affect response time.

It is switching power supply, but unregulated. 

All capacitors are inductive (even piece of straight wire is inductive), some less some more. Electrolytic caps are in "much more" category.  ESI (Equivalent Series Inductance) can be calculated.

Capacitor becomes self-resonant at the frequency at which capacitive and inductive reactances are even - usually <100kHz (very low) for electrolytic caps.  At this Self Resonant Frequency reactances are exactly in opposite phases canceling each other and capacitor is pure ESR resistance.  Above resonance capacitor is more of an inductor then capacitor.

Amplitude of charging spikes is limited by ESR of capacitor, transformer losses and resistance in series (fuse, cable etc).  Lowering ESR of capacitor might increase amplitude of charging spikes beyond maximum current of the rectifier or max ripple current of capacitor.  These things have to calculated (and not just tried), otherwise it is "garage operation".

Well, I guess that I'm a garage operator.... I have always made sure that my rectifiers would handle the current

I've been in garage operation when I was 16, building EL34 100W amplifiers with components I could find around including chassis made of heavy aluminum trays that fish store nearby tossed out, but when it became my profession for the last 40 years I've learned not to do anything that I don't fully understand.  I would like to understand how do you know what peak current thru rectifier is.  Did you calculate it and how, measured it or just used rectifier that seems to be working? Did you consider max ambient temperature or max line voltage?  Did you consider shorted output condition?

Czarivey, there might be individuals who are capable of such Toyota modifications, but I place them in "exceptional talent" category. Friend of my coworker did not like how transmission in his car operated, so he designed and built his own. Certain "arrogance" and great skill are required for that. Rest of us should enjoy what we have, especially when it is really nice piece of high end audio system. There is nothing wrong with caps that match closely original ones.

Timlub, thanks, I might take you on this offer one day - speaker world is foreign territory to me.  As far as amps go, there is nothing wrong in experimenting, especially if you enjoy it and have guru friend.