I would imagine some of that is due to websites having to serve multiple devices, screen sizes, etc. Back in the day, you just had to deal with monitor sizes.
To be fair, resolution is still the primary driver for "responsive" Web sites. Sure, we deal with more aspect ratios now, but you'll still often find yourself as a desktop PC user looking at the top portion of a banner designed for mobile devices when you first visit a page, mostly due to the increasing screen resolutions of larger phones and tablets. Ironically, the earlier versions of HTML were arguably better suited to the way we need information to appear on screens, as standards weren't especially strict and often came only with recommendations as to how certain items might be rendered. Basically, you'd rely on the browser rendering elements based on the device.
"HTML" is less so HTML now and more so a hacked-together collection of HTML, CSS and JavaScript, often obfuscated and/or overly reliant on frameworks that're supposed to make using them more tolerable. Aside from it generally being unintuitive, you're also expected to seek out information on proper etiquette when it comes to building sites, which in turn encourages you to misuse some of these technologies to add structure to sites where the more numerous tags of older versions of HTML might've made more sense. If HTML simply focused on creating structure for pages and showing relationships between content, then we wouldn't find ourselves having to design around at least 3-5 different device configurations (which, for the record, you currently have to do, even if you're using something to "simplify" the process, such as SquareSpace).
Basically, since it's a complete pain in the ass to do anything that doesn't fit squarely into one of the aforementioned frameworks, you end up with a lot of compromises in terms of how you can lay out a Web page without it breaking on another device or just becoming a chore to navigate. That's why content is so similar across the Web. It's virtually impossible to create a WYSIWYG editor for Web content with any real flexibility that doesn't rely on the end user understanding at least some code, due to how things like floats work, and people who want to avoid wading through the mess that is Web development by hiring a "Web developer" will often end up with someone whose idea of Web development involves tweaking a WordPress installation. To make matters worse, people get overly fancy in their implementation of certain features, leading to sites that're damn-near unusable in any predictable fashion, let alone usable at all by people with disabilities.
It's a sad state of affairs. I used to love Web programming, but I feel as though I went from having to do the occasional hacky thing to make something unique work to instead doing hacky things all the time to get basic, popular features on major Web sites working even remotely well. I'm always impressed when I see a complex Web app running, since I realize it must've been a complete pain in the ass to get it to work even that well, but when the big guys like Google and Microsoft can't even build a proper Web-based text editor that can reliably even open large files, you know it's a case of someone just forcing a square peg into a round hole and being okay with the idea that it's not gonna work 100% of the time. That's kind of the problem with the Web (and many Web services) we use now is: it's all basically just a less reliable version of something we used to natively install on our devices, or it's a Web wrapper.