Handwritten HTML with limited tags works just as well for many purposes (just forbid div, span, and a few others and the complexity you see in most webpages evaporates). The important part is using a text-based format from which information can be extracted even if the fancier display protocols become obsolete.
There are two different types of ownership here, and you’re conflating them.
One is the ownership of a digital copy on the same terms as a physical copy. That allows you to resell your copy, lend it to a friend, move it to a different device, retain the use of it even if the seller no longer exists . . . stuff that falls under the first-sale doctrine and other actions that are generally accepted as “okay” and reasonable. That’s what’s being called out here as not existing for most digital copies.
The other is the ownership of the copyright and permissions to reproduce additional copies. However, that isn’t what most people expect to get when they’re purchasing a copy of a media work, regardless of whether it’s digital or physical. How IP in general and copyright in particular is handled does really need an overhaul, but that isn’t a problem specific to the digital world—it’s equally applicable to print books, oil paintings, and vinyl records.
And to be honest, I’d prefer to see “lease” lose its meaning than “buy” go the same way, because apparently we can’t have both.
For some things, you can get non-DRM downloadable files, and those you do own. They’re very much the minority, though, and mostly limited to smaller, less-popular shops where they do exist.
I would very much like a law that says that streaming services and DRM’d downloads are required to use words like “rent” or “lease”, never “buy” or any synonym thereof.
Not really. HTML has a formal standard and definition that covers how to properly handle most corner cases that can arise when displaying it. Markdown has no overarching formal standard and exists in multiple dialects which are not always compatible with each other.
On the gripping hand, HTML involves more keystrokes (and technically speaking you need to include a bit of boilerplate in the file for it to be proper HTML). So it depends on whether you’re willing to do a bit more typing to make sure that no one can possibly confuse your italics with boldface.