I've had at least as many WebApps I've used built with JSP or ASP et al that broke back via unrepeatable POSTs and server side state as I have with SPAs.
While I agree that the back button _should_ work in an SPA, it's common for it not to. And it's common for other functionality to break. It's not hard to fix that and, for a well engineered SPA, such things typically do work. However, it _is_ extra work and it _does_ get missed. As such, SPAs tend to have more issues than non-SPAs. As such, as a user, I tend to prefer non-SPA implementations unless there is a strong reason that the SPA makes more sense.
If it is more common for functionality to be broken in an SPA than in a non-SPA, then that most certainly _is_ a valid argument against SPAs (from a user perspective). It's not an argument that all SPAs are bad. Rather, that I prefer non-SPA implementations because I can generally assume that it will have fewer bugs/annoyances in using it.
Why not? "You have to reimplement a huge amount of current and future behaviors that just work for MPAs" is definitely an argument against SPAs. That's extra work, extra potential for bugs, extra potential for accessibility problems.
It sort-of is if bad developers are less likely to wreck havoc with MPAs. Also, I don’t think that back-button breakage automatically indicates incompetent developers (although I tend to swear at them).
Not only is it not implicitly SPA behavior, as you say, good back button behavior design is actually hard in a app setting. And you can’t generally “get it right” — no matter what you do, in a sufficiently complex scenario (which… that’s why you built an SPA, right?) users will disagree about what the back button should do!
Generalized anti-SPA sentiment is silly. Saying “bad implementations correspond with SPA” is true but also silly because there’s not as much to get wrong in non-app settings which admit a “traditional” approach. It’s like the xkcd observation about heat maps and population density.
