This is the problem with wanting to use AOS4.x series, you are going to have to pay serious amount of money for a system.
I think the real problem is not the money, it actually is the contradictory philosophy behind it. On the one hand we see the decision for a much better hardware architecture, but on the other hand, the unwillingness to consistently look forward. In the long run was just that in many ways a tactical mistake. So I think the X1000 would be particularly for potential programmers a much more interesting business field, if one were to focus solely on the actual hardware, rather than remain permanently backward compatible.