Optimal dynamic experiment design as a tool for accurate estimation of microbial growth temperatures