I hadn't heard of it, but it's straightforward enough. Johnny decimal seems to work best for communication between multiple people, it's like a lightweight way of having human-implementable URLs.

The thing with PARA is that information falls into different buckets according to their use, not their topic. You're in a completely different mindset when you're looking up old stuff for reference (archives), remembering that bash command you use once a month (resources), etc.

On a first impression / hunch, your decimal system seems to have the failure mode that your old finances live right next to your current finances. Thus, the size of your context / working window just grows and grows without bounds, becoming unwieldy over time. With PARA, the idea is that a project folder contains data if and only if you need it to work on that project.

A very low-cost way of trying it out is to just implement the 2nd A (archives). That's by far the most important.