I don't care about the number of votes. If an expert can assert it meets notability criteria it should be kept. We should try to get such criteria for as many types of articles as possible.
Tell the people on AFD. (And I dare you to quote actual policy.) They seriously argue that a consensus of the admittedly ignorant on a subject beats a dissenting actual no-foolin' expert.
Simple self-assertion of expertise carries very little weight in AfD. This is true whether arguing for exclusion OR inclusion.
If, however, one has actual knowledge of a topic it is quite possible to influence an AfD _strongly,_ not by asserting knowledgeability but by _using_ one's knowledge to locate and cite convincing evidence weighing on the topic. One of the characteristics of Wikipedia is that the only authority one has is one's ability to convince other Wikipedians. This is sometimes a strength and sometimes a weakness.
The problem with saying that "If an actual no-foolin' expert asserts that an article meets notability criteria it should be kept" (or the reverse) is that on Wikipedia, there are no accepted credentials for actual no-foolin' expertise.
The current culture is that Wikipedians credit you with the exact amount of expertise you actually demonstrate _right now_ on the spot, in their presence.