Nick, Great discussion. My personal take is that we will see a "flight to quality" in the Web 2.0 space. I believe we're already seeing that in fact, with VCs starting to let some companies fail.
There are some fundamental differences from the late 90's: - Cost of Web 2.0 is quite a bit cheaper than Web 1.0. With the free LAMP stack (do you know any startups that are buying Sun hardware, Oracle databases, or iplanet web servers?), the cost of software infrastructure has come down by an order of magnitude. - With offshoring, the cost of software development has come down by 50%. - With the rise of standards on the web and many free web services, the cost of development a useful item, whether it's a product or a feature, has also dropped substantially. - At the same time, there are many more ways to see if a prototype can quickly get to scale. You have the likes of facebook, myspace, and other platforms that let companies develop a prototype for hundreds of thousands of dollars or less and see if it can gain millions of users.
As a result of all of this, VCs can de-risk a venture investment in a lot less time and for a lot less money than in the Web 1.0 cycle. This means a broader portfolio of investments and more experimentation, with the ability to focus Series B/C/D investments on the ones that show real traction, promise, and revenue.