Maybe I've missed it in the mass of Bonds posts, but there seem to be a couple of factors that aren't being brought up. I'm not sure how to fit them into all the other data, but they strike me as relevant.
The first is that it's not really very fair to compare Bonds AB/HR ratios with other players. Better to compare his PA/HR. Given the number of walks he's getting, one would expect his home run rate per AB to go up simply for that reason alone. Using plate appearances/home run will probably be a better metric of comparison.
The other consideration is that he moved to a new ballpark in 2000. A ballpark that was designed for his swing. Here are some numbers to compare.
Steroids Comparison: pre- and post-1998
1993-1997 AB/HR: 12.65
1998-2004 AB/HR: 9.63
1998-2000; 2002-2004 AB/HR: 10.15
2001 AB/HR: 6.52
1993-1997 PA/HR: 16.01
1998-2004 PA/HR: 10.50
1998-2000; 2002-2004 PA/HR: 13.87
2001 PA/HR: 9.10
Ballpark Comparison: Candlestick vs. [phone company]park
1993-1999 AB/HR: 12.66
2000-2004 AB/HR: 8.41
2000; 2002-2004 AB/HR: 8.88
1993-1999 PA/HR: 15.95
2000-2004 PA/HR: 12.14
2000; 2002-2004 PA/HR: 12.91
Clearly Bonds saw a huge jump in his numbers. The huge jump, though, correlates more closely with his move to [phone company]park than it does to his supposed steroid use.
I'm not trying to argue that he never used steroids. I think it's probable that he, and many others, used all sorts of illegal substances. I'm not convinced, though, that using steroids is as effective as detractors say. I think having a ballpark custom built for your swing is probably far more effective.
(And yes, I'm well aware he only plays half his games at home ...)
"The game has a cleanness. If you do a good job, the numbers say so. You don't have to ask anyone or play politics. You don't have to wait for the reviews." - Sandy Koufax