Proportion of GDP spent on research is not a 'meaningless' number. The EU spent a great deal of time trying to (unsuccessfully) urging its member states to push their total spending up to 3%. They've since realized that no single metric can adequately measure a nation's capacity to innovate in science - but this measure is still part of a basket of metrics that it's perfectly reasonable to use to examine a country's commitment to science.
To be clear - there's little evidence that spending a lot of money science will get you a Google or a Genentech, but on the other hand, a fair bit of evidence that spending nothing will make it extremely unlikely. ie science spending is necessary but not sufficient.
We warn the reader in advance that the proof presented here depends on a clever but highly unmotivated trick. -- Howard Anton, "Elementary Linear Algebra"