Supposedly they're trying to quantify what percentage of all bitcoin transactions are not made with the explicit intention of transacting BTC. For example an ordinal transaction is made with the intent of inscribing on the Blockchain rather than simply getting BTC from one address to another.
I think both cause he labeled the financial and non financial on his graph. The financial probably refers to the normal data like sending btc and its movement like payments or withdrawals from exchanges. While the non financial are those like what you explained ordinal inscription and maybe other stuff like metadata or messages.
But I think the website authors are counting all transactions including OP_RETURN as "non-financial".
Does that make sense? Please explain your thought process if you could.
I'm just trying to understand
Regarding that OP return is the highest in the graph means all non financial are labeled in that because its doesnt really related to the mivement of bitcoin since these are data like messaging, indexes or probably little data such as timestamp.
Its really a nice experiment on the author sides. But since its his study we cant say his wrong but someone with tech background with bitcoin metadata can definitely disagree with it if found some flaws or inconsistency especially its published on a website.