[zamubafoo]

I can see what you are saying, but I think it helps in two ways:

1. it is another way to conceptualize an idea. For most purposes this might not be relevant, but who knows where a hard accounting problem might be resolved through the application of graph theory (or the inverse!).

2. it is another way to visualize flows. Not everyone is financially literate or numerically inclined, so instead of handing them a table of columns of numbers and having them reason about the flows numerically, it lets you represent the flows spatially which maybe easier. After all, not all tools are for professionals.

Additionally, while the cumulative history in one graph might be much, simply adding filters based on transaction date might provide non-obvious insight that other visualizations miss. I can see this probably being even more helpful by cross referencing other information such as location.

[btbuildem]

I'm a big fan of looking at usual problems through unusual lenses - while something normally easy is contrived, the different perspective might make a hard problem much more easily grokked.

I wish OP included a working project that lets one ingest a ledger, producing a graph. Perhaps this graph view might be used for "forensic" purposes, where a one-line entry buried deep would stand out as an odd node?

Still a hard sell with only one very basic example provided -- especially in a domain that's been around forever and where any advantage is critical. The DE ledger seems like a basic, simple tool, iterated into perfection over millennia. You'd think you could improve it, but good luck if you try.