I once mentioned F024A or GP(12,5) really needed a name, and David Eppstein later came up with the perfect name for it, the Nauru graph.
When a puzzle based on 30-60-90 triangles became popular, I called the pieces polydrafters after the drafter's triangle, and the name stuck.
Heegner numbers 1, 2, 3, 7, 11, 19, 43, 67, 163 were named years after Heegner's death because his proof was weird and disregarded, but it turned out to be correct. And maybe because too many things were named after Gauss.
In 1979, Howard Garns developed the puzzle Number Place, and the term was trademarked in English countries. It became popular in Japan, where a more Japanese word was trademarked. It then became popular worldwide as Sudoku, because that was the name not trademarked. (In Japan, it's known as Number Place).
Due to $\mathbb{Q}(\sqrt{-1})$ being Gaussian integers, $\mathbb{Q}(\sqrt{-3})$ being Eisenstein integers and $\mathbb{Q}(\sqrt{-7})$ being Kleinian integers, I've made an attempt at naming $\mathbb{Q}(\sqrt{-2})$ the Hippasus integers, because Hippasus was famously killed for proving the immeasuarability of $\sqrt{2}$. Whether that naming attempt will be successful remains to be seen.
What are other cases where things got good names long after the fact?
The term Euclidean algorithm does not seem to have been used before the 20th century, according to this site. In antiquity it was just called reciprocal subtraction.