That is a dangerous over-generalization.

Let's say -ferinstance- you have kick-ass, seriously best-at-what-it-does software that's only localized in Japanese, and a uniglot USian attempts to use it. They're gonna have a hard time. The software will very likely fail them, because they will be unable to understand anything they're being told by the software.

Does that make the software shit? No. No, it does not. The software is great, it "just" needs to be localized in order to be great for a wider swath of humanity.

From the perspective of the user, if the product fails for that user, its a shit product.

I don't mean the product itself is actually shit.

Like I say:

If its works well for a user. It's a great product. If its fails for a user. It's a shit product.

From the perspective of that user.

But we can learn from both the users who like it, and the users who hate it. But we cannot blame the problems the user has, solely on the user.

Honestly, even this is a hazardous over-generalization.

I've looked at gobs of software that failed to do what I needed it to do. Much of that software was good-to-great at what it was written to do, but it -like a lot of software- wasn't written to do what I needed it to do.

That doesn't make the software shit, that makes it unsuited for my needs. I am -and many others are- clever enough to distinguish between the two categories. :)