>Because you could encounter absolutely any organism and make the same argument.
That's a function of the explanation being an extremely good explanation. It rises to the top precisely because it has explanatory power all across nature without evident counter-example.
>There is no configuration of organism that would cause you to say "huh, I guess Survival doesn't depend on Fitness after all!"
This is where the argument falls apart. For starters, species go extinct all the time for reasons tied to their evolutionary trajectory. And there are species still living that unfortunately seem very imperfectly adapted to their constraints and likely to go extinct without a run of good luck or human intervention (e.g. pandas). We seem perfectly capable of recognizing when such species are "on the ropes". Additionally there are relative advantages we can clearly observe from animals in overlapping niches, and we can marvel at the effectiveness of adaptations in ways that don't involve circular assumptions (e.g. algae's capability for efficient growth is astonishing and without equal on the planet).
And, we could surely conceive of preposterous examples that defy expectations (e.g. the other commenter's example of mice with 100ft teeth).
It probably feels like it proves too much, because it's confirmed over and over again in nature everywhere, at all times. But in an alternate world where that wasn't the case, counter-examples would abound (such as the mouse with 100ft teeth). So re-iterating the core lesson about the role of natural selection is not just a circular assumption, it's the culmination of hard earned, accumulated evidence, ready at any moment to be falsified.
The honeybee is a perfect example, because the stinger does pose a real question about how we understand it's relation to fitness, and it requires delving into all kinds of complicated dynamics about genetically related drones are to the queen, the role of the sacrifice in supporting the hive and so-on. If we didn't have explanations like those, it would indeed pose a problem with explanations that presume fitness.
That's a real payoff from being alert to the need to have robust explanations; I don't think anyone is just saying "well it's fitness" and calling it a day so much as they're honoring the explanatory power of a well confirmed theory.