I ordered parts to build a new GPS-Module cable and they arrived in a few days. Wires appear to be 30 gauge with thin flexible silicone-type insulation. Pins were much better than original ones and didn't have any strange plating on them. Crimps also looked fine. I carefully removed the Dupont (RC Type) pins from included single-position plugs and installed new pins/wires in a good 4-pin-position black Dupont plug from spare-parts box. Connection seems to be 100% now and not susceptible to drop-outs during basic flexing of cable or vibrations. During back-yard testing, I was getting a fast GPS-3D-Lock and good sats/hdop numbers. I'm calling it fixed.
|New GPS-Module cable installed. Pic is good enough to see pin-outs.|
I replaced the foam tape to be similar as shown above. I used 4-5 layers to make it as high as possible. The ceramic antenna on top is just barely touching the top plastic body cover (when fully installed) and seems to get the best reception this way (for this entry-level GPS Module that means just "acceptable").
Never again will I assume a connection is good, just because connectors are securely fastened (not on this model-aircraft anyway).
Upon inspection of Compass-Module, found the same problems. Luckily I ordered some extra parts to build more cables. New cable will end-up being about 1.0 inch shorter (but still reaches with length to spare). Like before, still had to carefully clean tarnish and corrosion from cheap imitation PicoBlade sockets themselves, soldered to the boards at both ends. Then, I installed the new cable.
|New Magnetometer (Compass) cable|
As you can see from the above pic, the pin-outs (for the stock Nova compass) are:
Update 2015-01: I have not had a single problem with GPS-Module or Compass since replacing these cables and cleaning away corrosion and tarnish (from the modules themselves).