a standard story for reuse may remain a futile goal, the constant need/desire to re-invent is infinite as is the domain
one way forward would be to universalise catagories or types of widgets in a common/universal framework and compete them
this way reinvention becomes a far more worthwhile goal than reuse, and reuse itself would be better served, and more likely to occur, by having a bigger/better pool of widgets from which to choose from
also at present i suspect polymer components have significantly more overhead than an iframe, since that is their import vector and they require many more files besides, but how many times larger, slower and harder to understand? (but overall that won't matter and the overall convenience win, will be well worth the highly inconvenient initial and complicated task of getting them in there).