Each simple datum is subjected to random circumstances, only the accumulation of many of them can possibly throw a reasonable idea about the situation.
So, 4-3 for Webber is not significant, probably is more random than facts (the same if it was 4-3 for Vettel). If we, eventually, reach a 20-19 score for Webber, it would point that probably those two are having a pretty equal season.
Rosberg is now "owning" 6-1 to MSC, that's becoming to look a bit significant, not just random. That's the reason I'll try to include (in the future) the "level of statistical significance" in each team mates score. So one could see what level of "factness"(!!) a particular score shows, or, inversely, what could be just random circumstances . 6-1 would give a higher level of confidence in "Rosberg is doing better than MSC" than 4-3 in "Webber is doing better than Vettel".
The beauty of measuring raw facts (or numbers) is they can be subjected to random events but they are not to subjective opinions. The "bad" side is we have to wait to accumulate enough evidence to make "factual claims".
With these stats we cannot decide who has more "artistry" as a driver, but we could reasonable assert who is "performing" better, beyond good or bad luck.
I find fun and interesting to make theories about what driver is better or worse in just the first races, I do it every time, but these statistics have a different goal: searching a long term "truth" (note the semi-ironic quotes
).