What causes data on the Web heterogeneous?

Is heterogeneity the main obstacle for data integration on the Web?
We may have the illusion that if data are described uniformly, then
automatic and accurate data interoperability is right on the way.
Many companies have well-defined data modeling
policy that enforces the uniform data model across the enterprise.
However, we still see a strong demand for enterprise-wide data
integration. What has gone wrong in an enterprise environment
with data quality assurance policy? Well, there are a number of
other issues involved in data semantics drifting. One of most
important issues is under documentation.

This is a photo for my profile 

Semantic Web or Web 2.0?

I have been doing research on the semantic web for a while. Recently, I came across the word web 2.0. It seems that there are more successful business models tagged as web 2.0 models. Although people in academic have put a lot of effort into the research on the semantic web, a real semantic web sit has not been emerged in any foreseeable future yet. What is the fate of the semantic web?