[...]I believe that this will provide many benefits. A minor one will be easier importing of transactions into financial and accounting software.
But metadata can already be recorded clientside. it provides the same benefits, without the blockchain bloat, plus room for additional data.
True, but it's more accessible and standardised if it's in the blockchain. True, the merchant provider, such as BitPay, can record this transaction category locally, and transmit the category information in the merchant protocol for the customer's software to pick up, but the history of the data would only be limited to the merchant, and only in the format supported by the merchant processor (if you use something other than Quickbooks, BitPay can't help you). Standardising this to be a part of the transaction would allow any client to label transactions with categories, and give both merchants and customers access to their transaction history categories.
An enormous benefit will be the ability to see at extremely precise detail what is being bought and sold globally, which would give even more information to the market as a whole to make decisions on what production and services to increase, and which to exit.
It's hardly "extremely precise" since anyone can attach any label to any transaction.
That's true, but I believe if adopted globally, the faked labels will be minor noise compared to true data, especially since transactions will likely not be free and spamming will be costly. I believe real time access to such information on a global scale, being able to see what is selling when, for what price, and in what direction sales are moving, would improve business and pricing decisions enormously. Currently, you have to pay a large fortune for even a small market studdy that only focuses on small regions.