Well, he's right - but only in comparison - simply because Google has HUGE page views in search. comScore and Nielsen estimate about 5B / month US searches, but I think it's more like 10-12B / month. So let's look at the chart another way.
Here are the top 10 Google domains, and from Hitwise data, the share of traffic (page views) each sub-site gets (for the week ending July 15, 2006). I've added a column to show what the page views would be if we assume that Google.com search gets 3B views per week in the US.
Site | % of all Google views | Est. Page Views per Week | Avg. Views / second |
79.6% | 3,000,000,000 | 4960 | |
Google Image Search | 8.6% | 322,491,523 | 533 |
Google Mail | 6.0% | 224,915,233 | 373 |
Google News | 1.4% | 51,990,456 | 86 |
Google Maps | 0.9% | 35,413,789 | 59 |
Blogger | 0.6% | 22,227,804 | 37 |
Google Video Search | 0.5% | 17,706,894 | 29 |
Google Groups | 0.5% | 17,706,894 | 29 |
Froogle | 0.5% | 16,953,410 | 28 |
Orkut | 0.3% | 12,055,758 | 20 |
So those are some pretty healthy sized websites. Just for fun, I added the fourth column of average views per second...
My comment on Paul's post is that Google's next big thing (if it ever comes) won't grow incrementally, but will have usage growth like YouTube or MySpace. In other words, to be anything like the scale of Google search, it'll have to be revolutionary.
Not many companies create two revolutionary products, but I think Google could be an exception.
BTW, I originally created that table in Google Spreadsheets, but there's no easy way to embed a spreadsheet in a page, like there is with Google Maps.