The primary reason humans reuse previous inventions is that point of economy of effort. We have a limited amount of resources that we can invest in any one project. If we were capable of independently recalculating the optimum design for each project we work on, we would do better. In the worst case scenario, we find out that the same solution that applied to our last problem also solves the one we're currently working on. But fairly frequently, we would find that there are different optimum strategies for different problems, and in those cases, we would do better. However, the gains would be small enough, and the frequency of this occurring would be low enough, that it is not worth pursuing this strategy with limited resources. If we had infinite resources, the cost of pursuing this strategy would drop to zero, and thus the expected utility would be positive.

I do think it is plausible that the same strategy would be reused because that strategy is optimal, but only in certain conditions. Namely, both strategies would have to be optimized for the same constraints. ATP is a fine source of energy, but any organism that wished to use inorganic phosphate ions for any other purpose would likely be better served with a different energy molecule. But, cells which do employ phosphate ions for biosynthesis purposes still employ glycolysis (and electron transport methods) to produce ATP.

For a more blatant example, it seems unbelievably improbable that the optimum anatomical design would be so incredibly similar between small, tree-dwelling herbivores(monkeys), ground-dwelling omnivores (apes), and the rational beings created to love and worship their creator.