Somewhere there is an engineer that argued quite vehemently that there is no way the air speed sensors on an Airbus A330 could possibly all fail
There is/was no engineer that argued this. Instead the argument was, "if this happens, what can we do to improve safety in that event?" That failure mode was thought of, I have absolutely no doubt. Engineers thought it was covered, they may have been wrong about that but I'll discuss that later.
leading the engines to stall in mid-flight
An aircraft stalling does not involve the engines, it involves airflow over the wings. Do you have any knowledge of the topic at all? Nothing I've read indicates there was an engine failure on that flight.
The aircraft crashed because when readings became invalid, the computer automatically disconnected the autopilot / autothrottle (as it should have). The pilots then made control inputs that were inappropriate for the situation. They were probably confused by the relative lack of data they had, and the multitude of warnings a complete air data failure causes. The pilots then held a nose up attitude through multiple stall warnings, eventually entering a period of extremely high sink rate. The aircraft had pitched up in excess of 35 degrees through this period, and the pilots held full nose up control inputs through almost all of it. It was the exact opposite of what they should have been doing. The pilots held the stall all the way into the ocean, impacting the water while still in a nose up attitude of more than 16 degrees.
I know people like to get up in arms whenever a crash is blamed on pilot error, but it's pretty clear in this case that the pilot's actions were inappropriate and their inability to recover from the stall despite ample opportunity will almost certainly be listed as the main cause of the accident. There were many contributing factors, but the data suggests that the aircraft would have flown just fine if given proper stall recovery inputs.
What could the engineers have done better? Indicate in a more useful way what was going on and which instruments were reliable. The pilots should have been able to tell at a glance what they should pay attention to and what they should ignore. The avionics display design may not have been good enough for them to do that. The stall warning may have deactivated inappropriately based on the invalid speed, because the computer thought the aircraft was traveling too slow for the angle of attack indicators to function correctly. This failure mode should not exist in my opinion. Either the angle of attack indicator should function at lower speeds, or an alternate stall indication should be used instead. Or just keep the warning on, since the aircraft is quite obviously not in landing configuration. From what I read, they were probably assaulted with a whole host of failure warnings that were confusing and may have contributed to a panic reaction.
Also, pilot training needs to be improved in some areas, especially involving loss of pitot static data. There is no reason an airplane of any type should crash because of a clogged pitot tube. This should be drilled into pilots starting with the most basic beginning flight training. I know from experience the topic is not covered at that level, besides a couple questions that may appear on the knowledge test. In fact, if I had not actually had a pitot tube get clogged during my training, I would have never encountered the situation at all.
There's some fairly good discussion about the events of the flight here.