The fridge I modified to use as an air source heat pump must be well over 30 years old.
My SIL's father is an HVAC engineer and we have chatted about a/c and what are common faults, and the subject of compressors did come up.
Seems the lubricating oil for the pump is not separated from the refrigerant (if I understood what he was telling me), and so you are told to let a fridge etc stand for an hour or two after transporting it so that the oil can properly drain back into the pump, in case it's managed to get into the pipework.
I expect running it without the full complement of oil is likely to shorten the service life to some degree.
It's weird the conversation turned back around to refrigerators; as I was going to post that I've seen exactly that same "melted & metal solder balls" failure mode in numerous run caps on compressors. I don't imagine SIL's father would consider consumer-grade refrigerators to be in any way "proper" refrigeration... that was the general attitude towards them from my shop foreman at the ding & dent appliance store and the crusty old commercial refrigeration techs I worked with at Kalamazoo as a teenager. The latter often said if it didn't use ammonia, it wasn't refrigeration.
The thermodisc (a prehistoric cousin to the PTC polyfuse) does not work as well as many people assume; compressors fighting to run with a weak run cap will often get quite hot before it trips, and repeated cycles do seem to increase the threshold to such a point where they'll get blistery hot.
It was SOP at the ding & dent appliance store to test every run cap with a cap analyzer (Google
Annie Cap Analyzer for some right proper horror shows - these are a scary bit of kit even if used properly) for value & leakage as part of the prep procedure; I found an amazing number of toasted thermodiscs and leaky/toasted run caps on units that appeared to run perfectly well, aside from current draw just a bit on the high side of MFR specs. If the current draw didn't drop down considerably with replacement of the thermodisc and cap, we knew that unit was not long for this earth.
As for the letting it stand upright; that is because the compressor inside a sealed unit runs submerged in an oil bath inside that welded canister. In normal operation, that oil only circulates along with the refrigerant in a vapor state. When the unit is transported on its side, the oil will travel into passages meant to carry only vapor. If the compressor is started with liquid oil in those passages, the oil is drawn directly into the compressor and it hydrolocks. The tolerances on modern compressors are so tight that this can damage rings and valves. On a used unit, any wear-induced particles are supposed to ultimately dump into that oil bath and be trapped there. As with any such system, the fine particles will turn into a sludge due to settlement. If this sludge is part of the oil which makes it to a vapor passage, it can clog the orifice/expansion valve.
In an ideal world, there would be a dessicant/particle filter in the high side to trap all this; and on many there are simply because it greatly decreases the infant mortality rate and is good for the brand. But on many there are not. Another victim of the race to the bottom, I imagine.
We knew to look for this on the units we serviced; they were much more likely to respond well to a simple thermodisc/run cap replacement and be a good reliable unit after burn-in.
mnem