I have a thought I've been stewing on. For most of the history of computing, the ability to support or operate software was pretty well correlated with having the ability to produce it in the first place. And thus you could start with a reasonable assumption that software was more-or-less fit for purpose by the simple fact of its existence. You could assume that whatever faults it had could be and would be remediated, and you'd be right so often that the remainder didn't make a difference. This is complicated a little by the existence of spam, or viruses, or app store shovel ware, but you could control for those things.