Search engine + HTML defeats complexity

I’ve been thinking about the semantic web and I’ve been to many a meeting where discussions of interoperability, standards, and so on have dominated conversations.  I nod my head in agreement and think about how if everyone would just stick with and use standards, all this data would be connected and the Internet would be a far better place.  But there seems to be an opposing current from the likes of search engines such as Google.  Their principle seems to be to lower the need for tight structure and organization of data and just have superb indexing and searching.  Gmail, which pushes tags instead of folders, and lets you search email quickly instead, is an example of this notion.  Recently at eBiosphere, one of the attendees mentioned that he tried in vain to find specific biological data by visiting traditional sources of highly structured data, only to give up, type the phrase into Google and get the answer he needed embedded in an HTML page.  While it would be better if each of those biological projects were sharing today correctly according to semantic web standards, what if Google + HTML already provides some of these answers?  It’s an interesting space, and it’ll be interesting to see how players Wolfram Alpha start to fill it.