Academia and publication (having lived in both worlds) have no shortage of barriers. Our ecosystem tosses out metadata-rich, well-define, extensible markup and publishes rigid pdfs in their place. Open-access papers (the product of several years of research) never make it into an index through which to have any chance of discovery. Or worse, they fall victim to link-rot and become inaccessible entirely.
The first step to resolving these problems entails 3 parts: Creating a complete index of all open-access research papers A programatic interface (API) for accessing, and citing content within these papers (like w3c openannotations) And an "air lock" entry point (some protocol) which ensures all new research is funneled / registered into this distributed index
My experience is, there's a fine balance between starting a new initiative because it addresses an un/under-served problem, and (xkcd on standards proliferation) doing so because of disagreement with how others are attempting to address it. In the latter case (which is pervasive), I see a tremendous opportunity in.