Here is one summary of a lot of the research out there. I have read a number of the original articles referenced and they seem to be good science. I've also run farily large temp swings myself for many years without issue.
http://www.co2science.org/subject/c/...chresptemp.php
:hmm2: Compared to a lot of what the Idsos have written about adaptability in corals, yes this is a pretty good article in that they get the science mostly right and there are only a handful of facepalm-inducing claims.
However, very little of what's in that article is relevant to hobbyists. Most of the mechanisms that account for the responses they're talking about don't occur in closed systems and those that do, occur only as a result of exposure to stress, which fluctuations (as opposed to spikes) aren't known to induce.
The example they spend the most time on- the Cocos Island recovery (which they neglect to mention was NOT statistically significant) was most likely due to selective mortality leaving behind the most dominant corals and killing off the rest. That's not really something you want to do in a tank.
P. lobata, the coral that made up the majority of the coral cover there and which suffered the least mortality, is also probably
the most heat-tolerant species of coral. There are documented cases of it living in water in excess of 100 degrees F, which can't be said for very many other species. It's not clear that its ability to survive extreme temperatures tells us much about any other species.
Another main mechanism that gets mention is symbiont shuffling. Basically when corals bleach, they can repopulate their zoox by taking some from the population free-floating in the water. What's often seen is that corals will have one clade prior to bleaching but will take up another, more heat-tolerant clade after recovery. Again, this probably isn't relevant to hobbyists because a) in most locations, these heat-tolerant zoox tend to be fairly uncommon except shortly after bleaching events, so most tanks have few if any corals that would act as sources for them and b) live phytoplankton doesn't last long in a reef tank, so even if you did have a source of clade D, it's unlikely that there are enough live, free-floating cells in the water to inoculate your bleached corals. It's also worth noting that these new symbionts tend not to form stable relationships with the corals. They are usually replaced again by the original clade of zoox unless temperature stress continues. It has also been shown that corals harboring these heat-tolerant clades may be more resistant to bleaching, but less resistant to bacterial disease, so it seems that rather than an adaptive solution to heat-stress, symbiont shuffling is more of an emergency response. This mechanism is also very limited as far as potential for protection from future temp increases. For corals that already have clade C (the most common clade), clade D is the only one that's more heat-tolerant. After you take up clade D (which some corals already have) you're already at the limit of adaptation via symbiont shuffling.
The production of heat shock proteins, which gets a mention, does occur in captivity, but it only occurs following acute temp stress- aka temp spikes. Non-stressful events like temp fluctuations below the mean maximum temp won't induce their production.