Here is the link to the free upload on deCODEme for 23andMe customers :
http://www.decodeme.com/data-upload
I tried it a few days ago. It's interesting, but I personally prefer the way 23andMe is presented.
In the Health section I find it annoying to have to agree to the terms and take a pretty useless survey (at least for young people who aren't likely to have had any of the diseases listed yet) before viewing each condition's results. Oddly enough, deCODEme gives me a low risk for 3 out of 4 conditions for which I had an elevated risk in 23andMe's clinical results.
I am disappointed that either company doesn't give more European populations with which to compare ancestry. Orcadians and Sardinians may be interesting for their isolation, but it would have been much more useful with at least 10 more regional populations.
Nasturtium, I wouldn't read too much in the global similarity. If you are of Welsh/German/Irish/British ancestry, as none of these population is tested, you can't know on which side you have inherited more. Closest to Orcadian doesn't mean much as they are of mixed Norwegian and ancient Scottish ancestry. I am closest to the French, the Icelandic, but it's inevitable as they have no sample population in between France and Scandinavia. I am pretty sure that if they had had Denmark or the Netherlands or West Germany, these would have been my closest matches. Then what does the French sample represent ? Someone from Brittany, Champagne, Berry, Auvergne, Provence ? These are all completely different ethnic groups. France has more regional genetic diversity than the whole of Scandinavia + northern Germany and the Netherlands combined.
DeCODEme gives a small percentage of Asian and African admixture to all Europeans, while 23andMe doesn't (only a bit of Asian for Finns or Ukrainians). I first thought that deCODEme's analysis was more detailed, until I realised that East Asians (e.g. Japanese) also have a few per cents of European and African admixture. It is meaningless for a Japanese person, as the country has not had any immigration for the last 2500 years, and it was from Korea. In other words, if there is any segment with European or African similarity, it is purely coincidental or dates back from the height of the Ice Age (but then we are all East African or Middle Eastern, so it is pointless and misleading not to give 100% Asian ancestry).
I was also dissatisfied that deCODEme does not give haplogroup deep subclades. I also don't like their subgroup numbers that do not match any nomenclature. The SNP are not even indicated with the subclade, so we have no idea what that represents. That should be one of the most basic feature in a DNA test.
Otherwise it's not so different from 23andMe.