Google Polymer plans to bring frontend web development to a new level. It leverages HTML5 custom components. In HTML5 new new elements were introduced, i.e. <article> or <canvas> which were both not available before. With Google Polymer, such custom elements should become an entire library of specific elements that Google either open-sources or that can be developed by anyone.
The problem lies with the part where Google plans to 'offload' functionality onto the browser. These functions, e.g. Shadow DOM, HTML imports, Pointer events, need standardizing so that all browser support those equally. This process that is dealt by W3C however has been slow, forcing Polymer projects to rely on so called 'Polyfills'. That's essentially an "emulator" to mimic the behaviour a modern browser should do natively, except it's really slow across the board.
Without a widespread adoption of many of the forward-thinking standards across various browsers there is less incentive for developers to embark on Polymer as the backbone of new applications.
Google hasn't given up on Polymer as they are now in the process of launching the third version which partially rolls back some of the more outlandish features, while improving on those the W3C consortium could agree on. This is a good signal and means that Google realized it's better to wait for the W3C consortium to make progress together, albeit slowly. The alternative would have been to make Google Chrome the only compatible browser which could risk developers abandoning the Polymer architecture.
Yes, we use Polymer v2 and have open-sourced a long list of Web Components called TangereJS that are specifically tailored to business application use cases.