Update (2/4/13) via Richard Carson:
The main reason for this is that I gave up trying to get "all" of the CV papers and studies after 2000. This is clearly stated on p. 21 of the book. Between 2000 and 2007, the bibliography contains a sizable but decreasing fraction of the published papers and papers presented at major conferences.
It is also the case though that the Google Scholar numbers are somewhat too high. That is because Google Scholar includes papers that only tangentially mention CV and because Google Scholar will include papers on the "valuation of contingent claims", which seems to have become a popular topic in finance. The CV bibliography tried hard to screen these out. It is also worth pointing out that the date feature of Google Scholar tends to report the "latest" date when it tries to merge various version. Many publishers put up electronic versions of previously published articles after 2000. This masked to some degree the fact that bibliography is missing some of the relevant literature before 2000. It is almost complete in early years but as one approached 2000, the literature outside the U.S. was both growing rapidly and hard to track down.