Sorry that the discussion got more general without getting specific first!
I do not recommend any component rated for less than 85 degrees C, inside any of our amps (except of course the Quickie!).
That's because I am a moderately cautious engineer by trade, of course - I have not attempted to measure the temperature directly inside any of the cases, mostly because it will be so dependent on the circulation (inside and out), ambient air temperature, and power line voltage. Empirically, we have not seen failures with parts rated 85C or greater. We have seen failures of for example the original Jupiter beeswax caps in the S.E.X. amp - I can't remember but I think they may have been rated 50C? (Jupiter has since revised their construction, and we have added ventilation slots to that amp.) We've seen failures in C4S installations, so I've revised my design rules - many products now have heat sinks and/or use larger transistors.
For what it's worth, in the case of power transformers and semiconductors, I normally design for a nominal internal temperature of 100C, while the parts we use are rated for 150C. For the semiconductors I assume the under-chassis temperature is 60C - but I haven't verified that.
Also for what it's worth, the effective environmental temperature is usually around a 50/50 mix of sir temperature and radiation temperature averaged over the usual 4-pi steradians. If a part is close to a power resistor (for instance) it can see some high radiation temperatures!