I wonder what the relative likelihood of "natural but related" and "natural but unrelated" really is. If the probability of "natural but related" is as high as 1 in 500, it may after all compare favorably with the alternative.
It seems to me that your incredibly poor probability estimates stem from a complete unfamiliarity with even the basics of orbital mechanics. If I had to come up with a number for "natural but related", it'd be orders of magnitude less probable than that.
This was my thought process: Let's suppose that a "closest asteroid observed" record happens once every few years. Now suppose that a once-in-a-century fireball is going to happen. What are the odds that it will happen on the same day as a "closest asteroid observed" event, assuming that the latter are independently distributed with respect to fireball events? About 1 in 1000 (order of magnitude). And then I chose 1 in 500 as a representative probability that is greater than 1 in 1000, that's all - it didn't derive from any reasoning ab...
If it's worth saying, but not worth its own post, even in Discussion, it goes here.