I felt the review was quite professional and nicely organized, though we've all just come from many amateur/messy NDA-period reviews, and there's no such thing as a perfect overview.
I don't fully agree with the conclusions, but the P2 is all around a decent product when you factor in the substantial possibilities of upgrade. It's not outrageously good or bad at performance, pricing, power consumption, and usability/technical features. Lacking extremes makes it decent and close enough to competitive, but no more or less for the DT market. Of course, if you compare it with the original Phenom, which was really a flop, things look kind of rosy.
That said, here are the factors I consider as weakening the conclusions for this article:
When they compare system pricing, they consider DDR2 for the Core 2 Quads and Phenoms, and DDR3 just for the i7. But when they benchmark, just like in Anand's review, all the Core 2 Quads are running off DDR3. Well, when you're in the store, you're trying to compare the performance/value of Phenom II/DDR2 and C2Q/DDR2, and their benchmark may not accurately tell you that. Conversely, I could be wrong and DDR2-1066 CL5 and DDR3-1333 CL7 may be virtually identical for C2Q, but a control test over a subset of benches would have been prudent (or a link to a relevant article, if it's been done).
In the system power comparison, the only representative for the Penryn family is the QX9770, making it seem that Penryns consume more power than the Q6600. Ever since the first review I saw of the QX9770 (at Anand's), it's been apparent that Intel did something to raise power consumption of this processor. All the other Penryns consume much less power at the same Vcores and frequencies. This unrepresentative Penryn carried into the conclusion that the PII leads in system performance/watt (with the i7 left out because it's not in the cost range). But they produced no performance/watt table and did not even reveal typical Penryn power consumption for this mainboard/RAM setup. Additionally, the X48 mainboard is among the highest power consuming chipsets available (supports 32-lane Xfire), while the 790GX is a good wattage stepdown from the 790FX (GX supports only 16-lane Xfire, power figures referenced at http://hothardware.com/Articles/AMD-790GX-Chipset-Platform-Launch/?page=9). Had they compared boards for the same market (ideally, P45/790GX) and used a more typical Penryn, I don't think PII would lead in energy efficiency, though it wouldn't be so far behind like the Phenom I as to call outright uncompetitive.
I would be quite in agreement with recommending PII as an upgrade, especially considering how many are running A64's on eligible boards, and in light of how uncompetitive PI really was. The upgrade for PI is not necessarily urgent, though, and in a short while, a better PII will come out which isn't EOL (i.e., a final upgrade, after which you must replace CPU/board/RAM at once).
I just don't feel the same about recommending a PII build from scratch, as the data points to its being slightly behind C2Q offerings in either price or raw performance or performance per watt, though not by margins to call it a big mistake. Both platforms are EOL, and forgetting that would be the big oversight.