My estimate (source: sounds good in my head) is you'd need a dozen or so browser experts working full time for years to build a browser capable of rendering most modern "web-app" style websites.

The core specs have a lot of integration tests (one of the shittier ones written by yours truly!), and most of the specs are pretty readable for experts (I hate the CSS Device Adaptation Module Level 1 spec though).

There's just a lot of it and a lot of subtle interactions which is where the time would go.

If you were foolish enough to set many millions of dollars on fire* to do this you'd end up with a browser lacking in key non-core-spec areas too. Off the top of my head: print layout, security, JIT performance, HTTP2 / HTTP3, general browser performance, UI polish, PDF rendering, mobile version, plugins, and DRM "support" (good luck getting the DRM gatekeepers to let you bundle that stuff with your browser). Add some more years for all of that.

* and/or smart enough to make it an open source project and convince people to do it for free, see the other commenter's link to Ladybird below