The web standards are enormously complex compared to that time. Not sure if today's tooling can compensate this.

The complexity of web standards isn't necessarily a bad thing. A lot of the "complexity" (as usually judged by the raw size of the spec) is just existing APIs evolving to be more rigorously defined. Before that level of specificity, browser implementations varied and engineers had to make certain decisions.

A more "complex" spec can sometimes make the job of the engineer simpler.

That's not the kind of complexity that is the problem though. The complexity you have to deal with is just more and more features being shoved into the spec at breakneck speed, or if they are not in the spec, it's just Google adding them to their engine and everyone starts using them anyway to the quasi-monopoly of Chromium-based browsers.

Are they? I feel like the real difference is an insistence on actually enforcing them. There were some really strong standards at the time of IE5, it was just the wild west in terms of what got built. And then you had Tantek Çelik quietly building all of them in IE5 on the Mac. Just a question of where to look.

May just be me being a romantic. Work shut off access to non-Chrome browsers an took away Incognito Mode. Which is fun for QA work.