Github as CV has been debated already. Your employer may place restrictions on your ability to contribute to or publish open source. You may already be working 40+ hours a week and have other interests besides writing more code. If I'm not especially familiar with a project, I have no way to evaluate whether your contributions are valuable; I'm using the project owner's judgement as a proxy for my own. Maybe the project manager is a chump and the project is a mess, so does someone contributing 100 commits to that project count as a positive or a negative?
Let's say you one-upped AngularJS with a new library. How would you compete with that sort of marketing muscle? Devs are no different from regular people: they don't want to change, they'll be averse to re-learning things, and they'll downgrade you for not being backed by Google.
However the fault probably lies within the wider academia culture and the lack of expertise that academics/institutions provide to those wishing to leave and go to industry. Some PhDs genuinely do enter for the purpose of having a research career only to realise that academia is not for them and vice versa..