But you conveniently overlook the very next sentence: Overall, Davis and Weller estimated that condoms provided an 85% reduction in HIV/AIDS transmission risk when infection rates were compared in always versus never users.
In other words, "always" users of condoms are exposed to 15% of the risk run by "never" users. Is that a chance you'd take for yourself?
That's miniscule.
Is it minuscule? How many sexually promiscuous men are there in Africa? Shall we say (conservatively) 50 million? And what's the duration of their sexual promiscuity? Let's be ridiculously conservative and say it's just 10 years. That's 500 million "person years". 0.9% of that is 4,500,000 instances of seroconversion. Now, with luck, not every instance of serocoversion is going to result in infection. But best case, we're still talking about hundreds of thousands of new AIDS cases over the next 10 years. You still like those odds?
And of course, as far as condom failure is concerned, "seroconversion" is just lagniappe. The real action's where it always was, in breakages and slippages. According to the very study you've cited:
Approximately 3% of couples who reported using condoms consistently and correctly (considered "perfect use") are estimated to experience an unintended pregnancy during the first year of use (123), based on results of one rigorous controlled trial as well as modeling based on rates of condom breakage and slippage. In a recent well-controlled randomized clinical trial of monogamous couples using latex male condoms for contraception over six months, the pregnancy rate during "typical use" was reported at 6.3%, with a 1.1% pregnancy rate during "consistent use" (45). Most of these couples had experience using condoms. However, based on estimates from National Surveys of Family Growth (123), 14% of couples are estimated to experience an unintended pregnancy during the first year of "typical" use, a failure rate that includes both inconsistent (non-use) and incorrect use, as well as breakage and slippage. Failure rates in the second year of typical use are about 50% lower (167).
Not such good odds, are they? Remember, these are pregnancy rates - as you must know, pregnancy is actually rather difficult to achieve, requiring near-perfect alignment of the stars. Mere transmission of semen is not enough for pregnancy - but more than enough to transmit the AIDS virus.
Your own study is a damning indictment of the view that condoms can save lives, and its implementation would be a death sentence for millions.