Not sure how valid those figures are, but even if they are accurate uniparental markers can be subject to founder effects that eventually skew proportions. They can be used as a proxy, but I think that's it, estimating autosomal proportions could be misleading.
I actually think one of the hints for Iberians having less Germanic ancestry than north Italians is what Ryukendo pointed out a few days ago:
You can see it in
other PCA plots too. The pull on Iberians seems to be weaker, which suggests - but nothing more than this - that influence was weaker here. Iberians are more homogenous, whereas north Italians are all over the place. Oddly Portuguese also appear to have some decent variation, given how small the country is, but we don't trend towards the French/British/etc.
I wish we could to the same with Global 25, but unfortunetly the regional Iberian references are too few to draw any conclusions.
Here's a close up of the Iberian cluster of G25 with all European populations included on the datase. Don't think much can be learned from it as far as Germanic ancestry is concearned.
Not sure David's
Celtic VS Germanic PCA helps much either