Are gas-phase models of interstellar chemistry tenable? The case of methanol