Derek, The capsule's pictures reveal corona oxidization. There was a period, In the mid-1970s, when a layer of the conductive gold sputtered onto the diaphragms tended to oxidize and form a corona-like halo, accelerated by the the electro-chemical reaction from the applied DC voltage. When corona oxidization is bad enough, the inner section of the sputtering can lose conductivity, so make sure that you test for increased steady-state noise when you encounter these capsules.
I am aware of the question - how could could gold corrode?- and confirmed the issue with Berlin: there was some kind of substrate material used to bond the gold to the Mylar. That material over time had an electro-chemical reaction causing the corona oxidization.
I've observed and dealt with the K47/67/87 corona defects since the late 1980s and have not found the issue getting worse with time; rather, the deterioration seems to arrest at one point.
I think for soldering this I might do it the poor mans way, heating a flat head screwdriver and using the tip as a soldering iron and gently attach the wires