I sat reading Fnord's calculations and I knew they were right, but they were not adding up to the same numbers as in Ken Warren's book. After doing some thinking, I see that Warren has worded something funny in his book that hides a few percent. So, after re-interpreting the text and then comparing to Fnord's math I found the few extra percent. The actual, real, honest to goodness numbers for a non-paired hand are:

The chance of flopping EXACTLY one pair by pairing a hole card: 26.939%
The chance of flopping EXACTLY two pair by pairing a hole card AND pairing on the board: 2.02% (this was the hidden 2%.)
The chance of flopping EXACTLY two pair by pairing EACH of your hole cards: 2.02%
The chance of flopping EXACTLY trips by flopping two cards to one hole card: 1.347%
The chance of flopping EXACTLY a full house, trips of one hole card and pairing the other: 0.092% (I left this out the first time.)
The chance of flopping EXACTLY four of a kind, three cards to one of your hole cards: 0.01%

That totals 32.428% as Fnord's math correctly predicted.

Source: Winner's Guide to Texas Hold'em Poker, Ken Warren; pg 197-198.