I've been having the "MK5 print starts, then stops extruding shortly into the job" problem as well, and believe I have now tried all the debugging steps mentioned above. After 10 to 20 minutes of printing, the extruder motor stops running, but everything else seems OK. The temperature of the extruder is OK, I don't think the extrusion stopped because it got too cool. And ReplicatorG (version 0019) shows no communication errors that I can see.
I've replaced the extruder board (it had another problem, needed to be done), moved the extruder and relay board as far from other components as possible, tried earlier versions of firmware and the most current version of firmware, put heat sinks on the motor driver chips, tried setting all extruder parameters to 0, and have ran with and without the heated build platform. Monitoring the motor voltage during raft extrusion with a multimeter showed a steady 11.7 to 11.9 V, watching it with a scope showed a fairly smooth 12V voltage. None of these has allowed me to get a print that will run to completion.
When I started using the MK5 extruder, it ran great, and I had many successful print jobs. I remember having the print stop problem, thought then that the motor was bad, swapped in the motor from the MK4 extruder, and started successfully printing again.
Now the problem has returned, causing me to try the debug steps mentioned above. I just ran a test extrusion (no steppers running, just a test extrude) that stopped after maybe 30 seconds. Before that happened, the sound from the gear motor started sounding more strained/slow/sickly, then the two lights on the extruder board clicked to one, and the motor stopped running. It's hard to hear the gear motor over the rest of MakerBot during a build, but I thought I could hear a difference in its sound from the beginning to the end of a build then as well.
Does this help with debugging at all? My current theory is that my gear motors worked great when they were young. But now that they're older, after they've worked for a 10 to 20 minutes, they do _something_ (pull too much current?) that trips the motor driver chips into a shutdown state.
I know this goes against the theory that electrical noise is causing the problem, but if it was noise, then I don't understand how my earlier prints were OK or how, after replacing the first MK5 gear motor with the MK4 one, I was able to get complete prints again for a while.
Is there more information I can provide that would be helpful? I really enjoyed it when MakerBot would print, and want to help find a solution to get him back on his feet again (so I can print feet for him)!
