Do genome-scale models need exact solvers or clearer standards? Article uri icon