I love <3 <3 <3 strange puzzles. This may be a bit stream-of-consciousness, but here it goes...
What I'm going on right now is that your board is fine and the problem is external. It worked in SA and all you really did was to change satellites and bumped your receiver around a little in travel.
My reasoning with the 50 Ohm load is to divide the problem in half with limited tools. If you see instability with no signal coming in, then the problem is internal to the receiver.
You are using an external active antenna. Obviously, you are connecting it to the inner of the two SMAs. I'm thinking that the LNA_IN port can be receiving some sort of spurious signal... Incidentally, I don't really know if their LNA_BYPASS connection can sustain a 50 Ohm load on it or how their receive switching circuitry works and how they shut off their own LNA/SAW filter when it's being used in the bypass mode. I suppose they may be measuring the bias current. If there is bias-tee current being drawn, then I would shut off the amplified path. Hmmm. Maybe what you are seeing is the receiver flipping back and forth?? I still don't know enough about how the operating theory of the DreamCatcher. Outernet seems to hold things very closely to their vest pocket.
OK ... Do you you happen to have an RTL-SDR dongle? Do you see a strong carrier pop-up every time your receiver goes nuts? You may need to make a strip-line filter of some sort (high-pass, low-pass, notch, etc.) to attenuate the interference?
If you happen to have a 1.5 Ghz sig-gen, you can see if the receiver is still stable under controlled conditions. I would feed it a nice CW signal at about -110 dBm and see if you get a nice, stable RSSI and SNR. Of course the signal state won't progress because CW does not have the right modulation on it, but it would tell you it's receiving a signal. But with the right lab equipment, I would now feed it a BPSK signal with the right data rate and symbol rate to see if it remained stable and go through successive signal states (to a 4) with a stable signal. If it's stable on bench equipmnent, then the problem is probably external to the board.