I am -NOT- purposing this as a mutually exclusive alternative. This should be an addition to what's already happening. No one who goes through the trouble of using this is going to say "well she doesn't have HIV, so what the hell, no condom!"
If you're already practicing to be safe, you still worry about the possibility of something going wrong, and you contracting the infection anyway. With this, you can be more certain that nothing will go wrong. You would still use a condom just in case.
Your argument is like saying people engaging in random sexual activity should close their eyes because not seeing warts will lead them to believe the partner doesn't have warts - which may lower then chance they use a condom. This would offer more information, and if it turns out that person is HIV positive by the test, you'd be eliminating your chances of contraction altogether.
The problem is that there is a very real possibility that someone who would go to a bar and test their partner there would then proceed to have unprotected sex instead of using a condom. A test could give them a false sense of security.
While you personally might still use a condom, judging by the reaction to this in the comments on engadget, many would not, not realizing that they actually are putting themselves at greater risk.
When/if this makes this to the FDA, they're going to look at the overall public health risk. If they assess that such a test would lower condom usage, they'd be unlikely to approve it. If you want to market this as such, you'd have to prove that people using it are just as likely to use a condom, and I think you'll have difficulty with that.
The reality is that ELISA tests in the US are cheap and fast enough already; anyone who wants to test themselves faces little barrier. The cheapness of this test is important for countries where that is not the case.
If you're already practicing to be safe, you still worry about the possibility of something going wrong, and you contracting the infection anyway. With this, you can be more certain that nothing will go wrong. You would still use a condom just in case.
Your argument is like saying people engaging in random sexual activity should close their eyes because not seeing warts will lead them to believe the partner doesn't have warts - which may lower then chance they use a condom. This would offer more information, and if it turns out that person is HIV positive by the test, you'd be eliminating your chances of contraction altogether.