So Wayne - what's your backsweetening process?
For others... WayneB is a widely respected gotmead dot com member/mentor.
Uh, oh. I'm not incognito any more.
Well, I guess it's nice to be recognized.
Anyway, here's what I do when I'm going to backsweeten.
Basic backsweetening process:
1) Figure out how much you want to sweeten, by taste (see below).
2) On sweetening day, add enough metabisulfite to bring the free SO2 level up to around 100 ppm. (Again, this depends on pH and whether there's anything in the must that will bind sulfite -- or you can just rule-of-thumb it.) NOTE: This is higher than hightest's recommended 50 ppm level, but as I said earlier, SO2 will eventually dissipate with time (even in a corked bottle), so I rather err on the side of caution and include enough to be sure all yeast and/or ML bacteria are wiped out. If memory serves, commercial wineries are allowed to add significantly more than that (350 ppm), and the sensory threshold for most people is even higher than that.
3) Immediately add potassium sorbate; about 125-150 ppm, to provide effective yeast birth control. That's around 3/4 of a gram to 1 gram per gallon.
4) Add my backsweetening solution, and then airlock the carboy and allow the mead to re-clarify.
5) Once it is clear again, before transferring to a keg or bottles, add enough additional metabisulfite to bring the free SO2 level back up to 50 ppm. I do this by measuring free sulfite with a sulfite measurement kit, but you can wing it by assuming that nearly all the initial sulfite that you added has either been bound or has dissipated in the time you allow the mead to clear again.
6) Then bottle or keg it -- it will be stable at that time, and immune from any wild ML bacteria or yeast that may find their way in while you're bottling.
Anyway, my method to figure out how much to sweeten is pretty basic. As was noted in earlier postings in this thread, the actual perceived sweetness of any beverage is influenced a lot by what else is in there, so I don't try to hit a given SG target as much as I just sweeten to taste. That involves drawing off a small, measured volume sample of the mead, and then slowly adding a honey-water solution of known gravity (just enough water to get the honey to flow easily, around 1.300-ish net gravity), added in small measured increments, tasting the mix between each addition until I get the sweetness level that I'm satisfied with. If you always add the same amount of sweetening solution in between each tasting, and you count the number of additions up to the sweetness that you desire, then you can either figure out the net SG of the sweetened mix (or measure it directly if you have a big enough sample) and simply add enough honey to the main bulk of your batch to come up to that SG.
BTW - I've found that I'm much more pleased with a "backsweetening by mead" approach that I've done for my last couple of sweetening efforts, than I have been by adding unfermented honey. Because of the experimentation that I've been doing on ultra-high initial gravity fermentations (with starting gravities between 1.150 and 1.170), I happen to have a lot of dessert-sweet mead around the place these days. I follow the same basic approach as I described above for the honey-water additions, but I use a sweet mead instead of the honey-water solution. I end up essentially blending two meads, the dry with the really-sweet, until I get the level of sweetness I'm after in the finished product. I've found that the sweet mead additions will integrate and taste "finished" faster than the additions of honey alone.