Miscman,
That's actually a pretty tough question to answer! You're right, the GSH:GSSH containing buffer is slightly reducing, but that's actually a very positive thing for the refolding process. It would be entirely possible to use reagents that restored the S-S bond rapidly, but we don't want to do that, chances are the bonds formed are going to be mispairs (especially with larger proteins) and you'll end up with a big mess rather than the structure you actually want. By using refolding buffer, there is constant reversible exchange of S-S bonds ('Shuffling') which allows the protein in time to find its own native stable structure. It is this stability which is in competition with the reducing conditions, the former winning in the end...
Hope that sort of helps
R