I don't disagree in this specific case, but that's not always true.
I remember on a contract years ago another team working on a project - that started before I joined and was still going when I left - to build a big data analytics pipeline (hadoop, kafka, etc.) that achieved nothing. The whole thing could much more easily have been implemented with SQL Server (which they already had), PostgreSQL, or whatever.
At the very least it cost them a lot of wasted effort plus maybe a dozen sets of salaries and contractor rates (there was a mix). But it also stymied solutions in other areas because the answer was, "oh, we don't need to do that because the big data project has it covered." The opportunity cost was significant: the company lacked the capabilities it needed for as long as this white elephant continued lumbering on its way.
So yeah maybe if you only need to store the data and don't need operations (why even bother then), otherwise just use postgis.