Pretty thin. Even ignoring the accuracy of the source data, there are plenty of outliers on that graph. Before I'd assume vitamin D has any effect at all on COVID-19 death rate, I'd compare Sweden and Scotland.