Stall is defined in terms of EAS
Stall is mainly misunderstood to be defined by Airspeed...
Is compressilibilty a factor affecting stall speed?
Yes, especially when stall is defined by the buffet level at the cockpit
But compressibility is not the only factor influenced by high altitude / low air density. Handling characteristics, especially aerodynamic dampening is significantly reduced, which makes the recovery a bit more tricky. Test pilots may not realize, as they are used to it. Commercial pilots may be quite surprised by it.
The original question was:
"Are transport category aircraft stalled at high altitudes in certification testing?"
The answer is: yes, but not to the same extent as at lower altitude. Most stall issues are encountered when flying in the low speed regime, during approach and landing. A lot of stall testing is done with high lift devices in use, because this is when you get closest to the margins.
If simple stall tests at high altitude do not show any anormaly, you do not further concentrate on it. Differend general handling properties at high altitude with a higher risk of overcontrolling or PIO are not seen as a stall behavior anormally and not forther evaluated, as they are simply normal. But not trained... The same applies to the engines at high altitude, they do behave differently, but that is perfectly normal, well known and not a stall issue.
Actually additional buffeting at high altitude due to compressibility effects often helps to identify a stall earlier, making it generally "easier" to handle. And pilots should be less afraid to reduce pitch / lower the nose if there is 40.000 ft of free airspace below...