The language school I attended all but banned romanization. The idea was to learn, practice, and finally internalize kana and kanji as quickly as possible. Hepburn is just a band-aid when it comes to language study.
For people not interested in learning Japanese, however, a unified romanization could have its benefits. It just never struck me as particularly inconsistent to begin with, even after so many years living there.
I recently implemented an eagerly evaluated embedded Prolog dialect in Dart for my game applications. I used SWI documentation extensively to figure out what to implement.
But I think I had the most difficulty designing the interface between the logic code and Dart. I ended up with a way to add "Dart-defined relations", where you provide relations backed dynamically by your ECS or database. State stays in imperative land, rules stay in logic land.
Testing on Queens8, SWI is about 10,000 times faster than my implementation. It's a work of art! But it doesn't have the ease of use in my game dev context as a simple Dart library does.
I've lived and worked in Tokyo and Japan for the last few years.
In my experience, most Japanese companies offer salaries up to 10M-12M yen for the most experienced software engineers.
If you are aiming higher than this, you may have a small selection of companies available. In my personal experience, the only engineers I have seen earning 20M or more were at Google or Mercari.
You will undoubtedly have more options for work in your current situation.
For people not interested in learning Japanese, however, a unified romanization could have its benefits. It just never struck me as particularly inconsistent to begin with, even after so many years living there.