Remember, dewetting is not just an oxidation of the intermetallics or
oxidation of lead.  Cu3Sn is a very stable compond and when plently of tin
based solder is present will form about 30-40% of the intermetallic layers.
 The Cu3Sn intermetallic does not solder well even if it is not oxidized,
probably because it is so stable and not very reactive.  In a test that I am
familiar with, a piece of panel was plated with pure tin, reflowed and then
aged at 150 deg C for about a week in a vacuum oven (no oxygen) until all of
the intermetallic and plating was converted to Cu3Sn.  The test piece dewet
badly while the plated and reflowed piece wet well.  You can see the Cu3Sn in
the microsection if you use a ferric chloride based etchant and the
intermetallic is thick enough to see optically, it is gray layer next to the
copper in the silvery intermetallic layer.  I you have a SEM you can look at
it better.  Note: the the tin solder coating at the knee of the hole on a
board that has been plated, reflowed and baked will then generally exhibit
only Cu3Sn with a trace of lead.

Phil Hinton