So I've been debating about using server side code in my web pages to identify robots so that I can make my html mark up more expressive. Instead of having a plain jane 'div' tag where my header is, I would serve up the 'header' tag if the server detected that a bot was requesting. Let's assume that you guys wouldn't penalize me for having different tags served to the bot. So here are my questions:
1. Does google care about the semantic html5 tags yet, specifically for SERP, etc? If not, will you in the future?
2. If google does care about them, does it need to see that the document declaration is indeed html5, or will it implicitly recognize html5 tags?
3. Would you penalize the website for alternating tags based on bot or human?
4. If there are other considerations I'm leaving out, could you expound upon them for me?
5. Given that none of this would work, let's assume that I just use html5 tags without server stuff and basically make it impossible for anybody without a modern browser to view the pages, would html5 then be beneficial to SERP or other google-things?
6. If none of this is true why does html5 matter for semantics at all (I understand that html5 is useful for lots of other things, but it's supposed to expand htmls semantic range as well, isn't it)?
Any help you could give would be much appreciated.