Julia 1.9 brings more speed and convenience

Existing simulation and serious scientific code is mainly in Fortran, so projects that are tied closely to existing work may get off the ground faster without the overhead of switching to a new language.

For new projects, there's no advantage in using Fortran now over Julia, and the latter provides a multitude of benefits over Fortran, from interactive development to high level syntax to the ease of composing functions and datatypes from a myriad of public libraries.

Although calling Python from Julia is straightforward, if your project depends heavily on existing Python packages without good equivalents in Julia and does not demand high levels of performance, it may be more convenient to stick with Python.

Your strategy of programming in Python and writing the numerically intensive bits in Fortran or C++ leads to the “two language problem”. For most projects these hybrid programs can be replaced with a shorter, faster, and easier to maintain Julia program (unless, as noted above, you really need some Python package).