So do I tbh but nobody ever makes a proper business case in practise. If you are going to say that HTML is no longer the default protocol for serving content to browsers (and that it is therefore extra work to support rather than the other way round), it seems like quite a big step. I guess we can rationalise the tattered corpses of Java widgets, Flash, ActiveX, Silverlight by saying they had proprietary roots and no universal support but feels like there's a lot of handwavey stuff around client support, RESTfulness, SEO.