To really understand where the Tesla ranks in safety, we would need a detailed analysis of: the accident, how non-Tesla cars handled similar accidents, how the Tesla is designed to handle such an accident.
I do agree that we have no where near enough data to draw general conclusions about the safety of the Teslas, so I think it's inappropriate to say, "they are doing better[or worse] than the average."
Once the numbers of cars become large enough we can simply look at the outcome numbers, to determine how likely that outcome is. It doesn't matter if a car is "designed to handle such an accident". What matters is if it does handle a given kind of accident.
A Tesla car is unlikely to ever be involved in a fire at a gasoline station. We probably shouldn't give combustion vehicles a free pass there.
As more and more Tesla cars hit the road, confidence level increases in the stats that are derived. Early data is not meaningless; it can be indicative of later results.
I think this is precisely the idea they are combating. The initial news reports didn't have any details and just said there was Tesla on fire on the side of the highway.