Welcome to the forum.
The research results you are concerned about have been discussed on this forum a number of times. They are not typical of most studies, and the large majority of experts agree the rapid tests have almost the same performance as the laboratory-based antibody tests. (And I speak as a co-author of one of the studies you refer to, the one from Seattle.) The bigger problem with rapid tests is not missing people with HIV infection, but a somewhat greater chance of false positive results -- which can be very upsetting while awaiting follow-up tests to show the patient didn't have HIV after all.
1) Negative results with the current antibody tests, including rapid tests, are 100% reliable (or very close to it) by 6-8 weeks. However, if your exposure was particularly high risk -- e.g. unprotected anal with a male partner who is HIV positive or with unknown status -- a 3 month (12-13 week) test is recommended by many experts. If you proceed, you can definitely expect negative results. This time I suggest you have a laboratory based blood test, not a rapid test.
2) Here are two threads that explain our positions on HIV test reliability:
http://www.medhelp.org/posts/HIV-Prevention/-A-Question-on-Testing/show/1347755
http://www.medhelp.org/posts/HIV-Prevention/Need-your-help/show/1345664
3) Online testing services have their place, and in fact I recommend them strongly in some circumstances. (I'm well aware they serve only as agents for indepedent labs, like LabCorp or Quest.) Such testing is often a good option for someone who has been sexually active in general, but has no symptoms. However, if someone has a specific exposure they are concerned about, or if they have symptoms, it's often best to see a health professional in person. Professional advice by an expert can put all the information into context -- the nature of the exposure, the symptoms, etc -- and can better advice a) which tests to have and b) how to interpret the test results in light of all the available information. In addition, many online testing services take advantage of patients' anxieties by encouraging (or not discouraging) many tests that may not be necessary, often at great expense.
In your case, it seems you have an objective slant, and you're dealing with HIV testing only, not an extensive "STI panel" of multiple tests. I have no problem with it.
Regards-- HHH, MD