My test is usually just repeatability, not plausibility. It's fine if the ultimate explanation is "creatures with fly speeds can fly" so long as there is then a model of how flying works, and heuristics can still be developed, like "wings generally indicate flying ability."
Eventually you tend...