Well, okay, it's not a one-product company, but it's a one-revenue stream company, and, for our purposes, that's pretty much the same thing. One revenue-stream companies--especially one revenue-stream companies with a massive revenue base--are significantly riskier than those with a diversified revenue base. In this case, as goes search spending, so goes Google.
Yes, Google has a zillion cool products, many of which might one day generate meaningful revenue. The trouble is, as with Yahoo! in 1999 (immediately pre-disaster), these revenue-less products won't do a thing to stave off a train-wreck if/when search spending stumbles. The other trouble is that Google's search revenue stream is so huge that it will likely be years before another revenue stream provides some real diversification.
(Offering CPM-based ads is a good start toward diversification, but assuming the profitability for this service is similar to the regular AdWords profitability, this revenue stream will have to get massive before it makes much difference to the bottom line.).
Will search spending stumble? I don't know (nor, I suspect, does anyone). Most revenue streams do at some point. To act accordingly, however, one has to to know when that point will be, and I don't. Suffice it to say that anyone who wants a reminder about what happens to the revenue and profitability (and stock prices) of one-revenue stream companies when the revenue stream falters should check out Yahoo! from 1999 through 2002.
UPDATE: As two of the comments suggest, in the original version of this post, I mixed up some terminology (AdWords on non-Google sites / AdSense). Fixed now, I hope.