The good news is it generally isn't necessary to reverse engineer browser behavior when writing a browser. Since it's mostly fairly standardized, there's a decent test suite, and the major browsers are all open source.

Though this comes with some caveats:

• There are exceptions like the CSS viewport spec which was reverse engineered from an iphone.
• There are a lot of specifications because browsers have been around for decades and Chromium keeps implementing stuff, and it can be hard to find enough programmers to write all of them / catch up from a fresh start
• This is a somewhat unstable situation; if we lose even a single major browser engine it's easy to imagine Chrome maybe not bothering with standardization and just telling people to read the blog posts and code.
• Web pages will do nonsense like break themselves if you provide a User-Agent string they don't like. Mozilla has an ongoing compatibility effort where they sometimes have to override the UA string for specific pages. So less popular browsers are already playing from a disadvantageous position.