The difficulty that a hypothesis with very low prior probability (like this one) has is that in order for it to reach a level of probability necessary to compete with the more mundane explanations like dust or birds, it is going to have to be backed by very strong evidence. This is the Bayesian expression of "extraordinary claims require extraordinary evidence". The hypothesis needs to make some predictions that are exceptionally unlikely given those mundane hypotheses.

Unfortunately, it's difficult to see what predictions those might be. The authors of the paper don't give any, and their primary "evidence" for the hypothesis - that nobody else saw the object - is actually more likely given the birds/dust hypothesis than the comet hypothesis.

So yes, there's a tiny possibility that this could have been a comet, but if so, it's hard so imagine that it could ever be elevated to the point of a hypothesis that we should even take seriously, much less one that we should believe.