This really just comes down to what you're writing, how app-like your web app is. It's too easy to have one's own experience focused in a certain area and estimate the remaining majority as relatively similar. (For most of what I personally work on, the DB portion is mostly a simple straightforward serialization of what the user has built through the application; whereas client-side state has so many aspects to it I couldn't give a brief characterization—the whole app is basically client-side.)
From what I can tell most of the disagreement about SPAs results from devs who are building things that aren't app-like railing against their futility vs devs who are, who become perplexed by the vitriol when they have immediate experience with their architectural benefits.
The SPA criticics from the article and this thread have repeatedly said that their issue is not with building things that need the benefits an SPA architecture brings. The criticism is that the majority of SPAs are harmed by that architecture because it is the industry default and being used when it isn't appropriate.
Also, for what it's worth, I've worked full-time on one of the largest SPA projects in the world (~500 engineers contributing frontend code to it on an average week), so this is not coming from a place of total ignorance.
This was not my extrapolation. But if you skim over a comment you're likely to perceive it as falling into one broad camp or another whether that's the case or not.
This is a bit weird to me, in that I'd say that cuts in the opposite direction. State can exist in at least three locations for most applications: db, app server, and client. Keeping state consistent across all of them can be difficult in the best of times, but thin clients by their very nature carry less state, lessening the burden. Sometimes client state is necessary, for richer user interactions, but for all but the most cosmetic of purposes you're going to have to replicate that state on the backend anyway, to enforce business and security requirements.