You asked for comments on raising the profile of B4J, but I'd improve the prospective user's experience before you drive huge traffic to the site and potentially waste it.
My evaluation issues have been around unclear sample code and sample projects that don't work out of the box. People won't give you much evaluation time and will look to samples. I may just have been unlucky, and all your others may work perfectly.
A specific example is my unanswered question on using the "loginserver" sample project. It says down the very bottom of that project that two extra libraries are required but I cannot download them. The web page says I do not have permission. They appear to be part of a product I have to pay for. A sample project for a free product requiring things I have to pay for does not help a new user without either making it clear that that is the situation or explaining what to do if it's not. It's not the first time that question has been asked on the forum - the other one is as yet unanswered, too.
I've picked on one thing, but I'm sure if I or others were to look there'd be more. Documentation is a huge task, I know. But I think it makes the difference between something I'd pay for or not. I'm patient and persistent as I see the potential, but people in general make snap decisions and many a good idea has been lost to the big boys because people are put off at the early stages.
I know it's free (at the moment), but people will still evaluate it using their precious time.
I'm only writing all this because I want to love it. I think it's brilliant but damn, I need documentation, working samples or at least faster answers on the forum
Others will, too.