Is plain HTML really unsafe, or is this about scraping?

Plain HTML isn't inherently unsafe just because a scraper can read it. The likely issue is that a simple server-rendered page is relatively easy to collect at scale, while a JavaScript-heavy interface can add processing costs and opportunities for bot detection.

That raises the cost of harvesting rather than preventing it. It could push some collectors toward paid API access or discourage them altogether, but the company's actual motivation isn't established by that alone.

The headline sounds like a stretch to me. A change in access to an older interface is not the same thing as a declaration that HTML is dangerous. The reported access problem also doesn't seem to affect everyone in the same way, so I wouldn't assume the older interface has been removed entirely.

The frustrating part is the tradeoff for ordinary readers. More client-side code and access checks can make a site slower or harder to use, and may also make public discussions harder to archive. Those costs can fall on people who aren't scraping anything.

There may be other business reasons too. A site might prefer people to use an interface where it has more control over ad placement. That's speculation, though, and it doesn't rule out concerns about automated collection.

Don't assume every machine-readable route is gone. A JSON endpoint was reported to still work, though it may have rate limits. That's separate from whether the older interface works without signing in, and neither tells us much by itself about the site's intentions.