That demonstrates that strong worker protections doesn't necessarily decrease the number of businesses created. It does not demonstrate the original claim that they promote innovation.
If you look at how the data has changed over time as a proxy (in some cases) for uncertainty on the data, you can see that Sweden's current rates are new. In the 2000s, the US was well ahead of it. However, in the 2000s, the US had a Bush administration; Sweden's social welfare net hasn't significantly changed in relative scope during the last decades. If it is true that social welfare nets promote innovation and companies per capita are a good proxy for innovation, why has that changed?
Number of entities for innovation is a questionable proxy for reasons already mentioned in my grandparent comment. As another example, taking your example of France: regulation changes pretty drastically as company size increases, which promotes creating new companies instead of growing existing ones, artificially driving up the number of entities while not really having a meaningful effect on innovation and only having a small effect on de-facto corporate structure.
If there was a predictive effect, shouldn't we be seeing other evidence of entrepreneurship where at least some of these countries also do well?