Is a software glitch any more concerning than a rudder hardover? Or fuel icing flaming out both engines?
I would have to say yes. A software glitch can be much more difficult to prove/find/replicate, and is therefore harder to fix. It can also manifest itself in many different ways, making it difficult to develop procedures to deal with the problem before the fix is found & implemented.
It has also been my experience that engineers tend to dismiss these sort of problems as 'finger trouble' on the part of the pilot. If the machine & it's computers are designed not to do something, people find it hard to accept that it can actually do the very thing it was designed not to do.
A number of years ago I was flying the 767. In the cruise, when you changed the altitude in the MCP, the selected altitude would appear in the FMC scratchpad. This would make it easier to make the change to the cruise altitude in the FMC - just line select it in & execute. We were told that the system was designed to only send the signal one way - from the MCP to the FMC. A signal couldn't travel the other way, from the FMC to the MCP.
Well it happened to me not once, but twice. An altitude appeared in the scratch pad without any pilot input & then the MCP altitude changed to match. The first time I didn't immediately notice the MCP change, as I was looking at the FMC wondering how the altitude had magically appeared in the scratch pad. The second time it happened (different day, different airframe) I immediately looked at the MCP & witnessed the digits in the altitude window rolling to match the altitude in the FMC sratchpad.