The only prediction you made was that HTML5 will make significant in-roads against native UIs. The rest of what you wrote at the end were directives.

I think this prediction is premature. HTML5 has been hyped up to be a replacement for desktop UI, it is not even a solid standard yet. On top of that there is lag across numbers of browsers in supporting the functionality of HTML5, and that is due in part to the "standard" still being in flux.

Another thing standing in the way of this is the fact that you have to use Javascript and CSS to build out a UI. Okay, you need them and HTML5 is replacing the need for some Javascript, but the requirement for a mix of languages/formats just confounds development. So things are not really that much ahead of native UI where you can use code in the same language as the business logic or, depending on support/toolkits, have other formats mixed in (e.g., CSS, XML).

Also, there can be contenders that might still make desktop UI attractive. JavaFX is one. Still in its infancy in some respects, but quite promising if some missing functionality can come in within the next year or so. And it's support for web-based operation could be a real challenger to HTML5. JavaFX can even do HTML5, Javascript, and CSS with its Webview and Webengine components, in a desktop or web app. Mobile support is a being worked on.

My prediction is that HTML5 will take upwards of 5 years to make serious in-roads over native UI. And at that we can expect to see other technology to challenge it, which can incorporate it along with its own approaches into desktop and mobile beyond browsers, blurring the lines between web and native apps.