Sorry Claus but your analogies are ridiculous.
Firstly, if a meteor strikes and takes down a computer then that is not something that can be planned for by either the computer, the programmer or the user (the odds being so extremely low!). And i strongly doubt anyone would be blaming that on a computer error or a user error. The blame may go to the users for not having a good enough disaster recovery plan to regain the lost information in a quick and orderly fashion, but the strike itself would not be blamed on a computer error.
Secondly, within the earths atmosphere ionising particles are not a risk to computers, however, in space this is a definte risk and it is one that is taken into account on every satellite that is currently flying above this big blue sphere we call home. It is such a common problem that every program that run on computers in space must be designed so that should an ionising event occur (and there are a number of different types causing differing effects) the program will be able to detect as much and either reset or shut down the computer to prevent further damage. Therefore, if an event occurs in this way and does cause a problem then i would suggest that this is a failure of the programming to anticipate a known and likely event and to precipitate the correct instructions. So yes this is a user/programmer error.
On your last point "So what do we need Gods for then?", well the fact is that we as rational sentient beings do NOT need gods. But some people still feel more comfortable believing in a higher power/sky fairy/etc. That is life... It has no bearing on the differences between computer errors and user errors.
Finally to refute your argument i offer this - Can a knife be in error? if it cuts someone is that the knifes fault or the users? If the blade is dull perhaps that counts as an error? Or is that an error of the user for not sharpening the knife? Obviously a knife can be broken, but will a knife every break on its own? So that would be the users fault for using the knife incorrectly, no?
By the same logic is there any dumb object that can be said to make an error? I would say no, only the user of a dumb object can cause the error. And computers are no more then dumb objects, we attribute more to them then most things because they are so much more complex but all of that complexity is down to either programming or user control. So any errors i would suggest automatically fall into those categories...
Just a thought...