"you rightsize the CPU to the task in hand"
My problem is Calxeda chose a point workload that was right sized for their server and then said that the other server was not as good. To compare a workload that is tuned for your configuration and then to compare it with another system that has a different "tuned spot" is OK. They were just dishonest about computing the metrics to compare.
I kind of agree with you, but...
The whole point of these 'poor cpu' servers is they are only usable for certain workloads, so you wouldn't expect Calxeda to choose a workload that highlights how poor they are:)
The webserver example is a good use case for them, although as you point out an Atom server is a better comparison. Calxeda's numbers are for a full node (SoC, memory, network etc) so you'll need to add them to the Atom server.