Tom MacWright on a potential new start for the Web.
The document/application dichotomy I totally uphold as (close enough to) true. And the idea of a Markdown-based document web 2.0 I can get behind completely, as long as it's the CommonMark with table and definition list extensions. (Or is it Web 3.0 now?)
The eternal problem with such proposals is being able to gain enough traction. The web has gotten thus far mostly because it has, to a first approximation, never broken backwards compatibility. It'd be darn hard, if not impossible, to convince a meaningful majority of the rest of the Web to adopt a completely new proposal.
Just the scope of the Web makes this a hard problem to solve. But yet the scope creep isn't stopping any time soon.