> the choice of the P1022 was made due to availability at the time A-Eon's choices
> had to be made. Hindsight is always going to look better and sometimes, the
> choice made may not turn out to be the best available when looking backward.

I wouldn't say it is a hindsight 20/20 thing in this case. In hindsight, the choice of the PA6T for the X1000 probably wasn't the best one (bad availability, too expensive, weaker than expected), but it was the most sensible choice back when the decision was made. With the e500v2-based P1022 I can't think what came over them as the technical implications of attempting to run legacy code on this core had been discussed before (see e.g. UltimatePPC fiasco).

> It is always a gamble [...] when committing to purchase in quantities such as [...]
> 1,000 units for the [...] P1022 CPU's.

Purchasing a fully compatible Applied Micro CPU or waiting for sample availability of the fully compatible Freescale QorIQ T1 would have been much less of a gamble in my opinion.