I don't think so. If the Internet was all that was needed to explain this concentration, then that would not be a sufficient explanation because it would prompt a new question. If geographical proximity was due to the Internet, then why Silicon Valley vs some other location?

I think that it is more likely that there are a high concentration of tech startups in SV because the concentration of tech companies in SV offers positive externalities to firms that locate themselves in SV. This is not a very good article (even by wikipedia standards), but I think that it could help to paint a picture:

http://en.wikipedia.org/wiki/Economies_of_agglomeration

But then, you are right. If tech companies needed to be within a hundred miles of their customers in order to transact business with them, then they would not be able to benefit from agglomeration effects. But I think that there are probably other things that factor into as well - margins, competition, regulations, etc.

"Brick layers / construction are far less mobile because you need to be at the construction site to build a brick wall."

I agree, this is true. Perhaps I should have been more selective with my examples. One more thing that is also true is that a side effect of their lack of mobility is that it is more challenging for them to scale vis-a-vis the technology companies.

"PS: Read the millionaire next door and you find a lot of people in the use that made a few million from those bricklaying startups. The main difference is it often took them 20 years to get where software companies got in 5."

The OP cited bricklayers as an example of million-dollar startup companies not being unique to the computer industry. I think that your point about how challenging it is to scale more conventional "mobility-challenged" businesses is actually a very convincing argument against the idea that starting a computer company and starting a bricklayer company offer comparable potential rewards.

Holding all other factors constant, if bricklayer startups are more geographically limited than tech startups, then it is reasonable to hypothesize that tech startups have a higher probability of becoming million-dollar companies.

Also, with regards to the Millionaire Next Door, I think that it is worth repeating the old axiom: the plural of anecdote is not data.