Is that what this can be, a media center pc for only £200?

Can it also mean a laptop the size of a DVD player, like an EEE, but with the power of a proper graphics card and decent sized hard drive? Imagine a proper gaming laptop with a 9 inch screen for £450, now that's appealing to me.
The calculations used in the comparison of Atom Vs. Nano to find total energy used to complete the tasks are using the total system power multiplied by the amount of time each system took to complete the task. This is invalid, as even in the comment right below the power numbers the author notes that the total power should not be used since it is skewed by the use of a very large PSU which will be inefficient at these loads. If you only compare the additional energy used by each platform as it completes the task (465 seconds * 20 watts for Nano and 605 seconds * 4 watts for Atom), for the MP3 encode you actually get 9300 watt-seconds for Nano vs 2420 watt-seconds for Atom, making Atom 380% as efficient as Nano.
Ars Technica seem to have found a potentially explosive bug (or worse, deliberate fraud) with PCMark 2005. It looks like PCMark 2005 has optimized codepaths for AMD and Intel (they found our as the Nano lets you change the CPUID to pretend to be another chip).

It's buried in the middle of their review:

http://arstechnica.com/reviews/hardware/atom-nano-review.ars/6

This needs further digging!
Have seen a couple more Nano V Atom reviews, specifically [H]ardOCP's and the Atom gets thrashed. However, Atom is 5w tdp and Nano is about 25, so there is a big difference!
Is that what this can be, a media center pc for only £200?

Can it also mean a laptop the size of a DVD player, like an EEE, but with the power of a proper graphics card and decent sized hard drive? Imagine a proper gaming laptop with a 9 inch screen for £450, now that's appealing to me.
The calculations used in the comparison of Atom Vs. Nano to find total energy used to complete the tasks are using the total system power multiplied by the amount of time each system took to complete the task. This is invalid, as even in the comment right below the power numbers the author notes that the total power should not be used since it is skewed by the use of a very large PSU which will be inefficient at these loads. If you only compare the additional energy used by each platform as it completes the task (465 seconds * 20 watts for Nano and 605 seconds * 4 watts for Atom), for the MP3 encode you actually get 9300 watt-seconds for Nano vs 2420 watt-seconds for Atom, making Atom 380% as efficient as Nano.
Ars Technica seem to have found a potentially explosive bug (or worse, deliberate fraud) with PCMark 2005. It looks like PCMark 2005 has optimized codepaths for AMD and Intel (they found our as the Nano lets you change the CPUID to pretend to be another chip).

It's buried in the middle of their review:

http://arstechnica.com/reviews/hardware/atom-nano-review.ars/6

This needs further digging!
Have seen a couple more Nano V Atom reviews, specifically [H]ardOCP's and the Atom gets thrashed. However, Atom is 5w tdp and Nano is about 25, so there is a big difference!