I do not think it has really been established how high (or low) the salinity needs to be and for how long to kill any particular coral. Most folks do not like to try such experiments.
The bugger of it is that corals seem to be strongly specialized for the normal salinity of the reef they are on. Corals growing on a reef in the Philippines with a normal salinity of 32 ppt might experience real physiological stress if the salinity rises to 37 ppt whereas corals from the Red Sea, even of the SAME species, may live normally at a salinity of 40 ppt and not show those same signs of stress until 45 ppt.
What is completely unclear is if these responses are due to physiological or evolutionary/ecological adaptation. Can you slowly acclimate a Philippine coral to a salinity of 40 ppt and stave off physiological stress until 45 ppt after giving it, say, a few months to acclimate, or are these responses different because the different populations of corals (Philippines vs. Red Sea) have adapted to local conditions over evolutionary time? Are these responses due to acclimation or are they innate?
As far as I know, no one has ever done a study to look at that problem in depth. People HAVE looked at the salinity tolerance of corals within the context of local conditions, and as I said, tolerances vary geographically, even within a single species of coral.
On the other end of the spectrum, corals in the Philippines and other parts of SE Asia normally experience relatively low salinities as compared to most places (30-33 ppt is common). The corals grow and thrive in these conditions. On the other hand, corals from parts of the GBR are normally exposed to slightly high salinities--~37 ppt. Corals from the GBR, when exposed to salinities of 30-32 ppt, which are just fine for corals a few thousand miles north, have all kinds of physiological problems even experiencing bleaching within a few days. At salinities in the upper 20's, which corals to the north survive for weeks sometimes, these corals begin dying within hours.
The closest study I am aware of that gets toward any of this is the unpublished work of a friend here. She just defended here thesis, and I presume the work will be published within the year. Anyway, she looked at Siderastraea in Florida Bay. Different parts of the bay get different freshwater input and while some parts are usually at a salinity above 30 ppt, some parts will fall much lower rather frequently. In most of the populations she looked at she found similar behavior across the board, with the corals not really having toooo many problems until salinty fell to 10 ppt (these are hardy, hardy corals). In one population, however, and one that rarely sees salinity below 30 ppt for any length of time the corals experienced physiological stress at higher salinities (~15 ppt if I recall) than the others.
This is far, far, far from certain, but this may suggest that physiological acclimation to local conditions may be important and evolutionary/ecological differences between populations may not account for all of the variation. It would take a lot more work to figure that out, and that wasn't really her aim anyway.
So having said all of this, I think that yes, a salinity that high could be stressful for at least some corals, but may be fine for some others--impossible to say for certain. I would slowly drop salinity as you are doing and shoot for a salinity of 34-35 ppt most of the time. There isn't a coral or reef animal out there that won't thrive in that range and 1-2 ppt higher or lower either way over the short term should be quite tolerable.