I would expect a harp to go sharp as the humidity goes down; the soundboard
is shrinking and pulling the strings tighter. Other than that you may not get
much warning. I once had a hammered dulcimer crack at a show. The hall was
'way overheated, on a particularly cold, dry Thanksgiving weekend. The poor
instrument put up with as much as it could take, and then opened up a 1/8"
crack the full length of the hardwood top. It sounded as though somebody had
hit the thing with a hammer. Hard. I must have jumped a foot (.305 meters,
for those of you at the Gamla Fiskar Conference).
Several years ago I saw an article in 'Fine Woodworking' about shrinkage in
wood, related to frame-and-panel construction. The basic conclusion was that,
if the piece has nowhere to expand during high humidity periods, the long
term effect is toward shrinkage. In other words, in the expansion part of the
cycle, the wood takes up the stress by compressing, and then shrinks further
in the next drying cycle. So, it would seem that the thing to avoid is taking
the intrument through too many extreme cycles. Since most of the motion
occurs at the limits of humidity (if I remember right) it follows that the
best course is to build close to 50%, and then try to keep the instrument
between, say, 35% and 65%. Obviously, there would have to be some
accomodation for where you expect the thing to live, i.e. desert or swamp
conditions.