Indeed, the biggest advantage of paper arithmetic is that it leaves a written record: each step in the algorithm can be checked for mistakes afterward. It is otherwise significantly slower and not inherently more accurate than a counting board.
The other advantage that paper arithmetic has is that it can be easily reproduced in printed books, making learning more advanced techniques more accessible / easier to spread to anyone literate, without requiring an expert teacher.
(And finally, paper arithmetic [eventually] has the advantage that it can be more conveniently extended and generalized to include more kinds of operations and structures, in a way that is easier to explain and teach than adding new kinds of counting board rules. Paper arithmetic is a more natural precursor for symbolic algebra than counting-board calculation.)
The big disadvantage of paper arithmetic is that it depends on widespread literacy and cheap access to paper (or similar material). In a context where paper is expensive or unavailable, written arithmetic is not very compelling.
Greek and Roman students typically used wax-coated boards and scratched them with styli. The surface could be heated or rubbed to restore its unmarked form.
My point was that there were cheap alternatives to parchment and paper for quick work.
The thing that really spread Hindu numbers around the world was access to paper.
My impression is that in India “Hindu” numerals and arithmetic were developed using some kind of erasable board, but among Arabic speakers (and later Europeans) the spread of Hindu numerals went along with paper as a prerequisite technology.
This looks relevant https://cpb-us-e1.wpmucdn.com/sites.ucsc.edu/dist/f/704/file...
Here is a relevant looking book: https://yalebooks.yale.edu/book/9780300089554/paper-print
In e.g. Britain, the transition to Hindu numerals did not occur until 1600–1800, depending on the place.
It's hard to have double-entry bookkeeping without books.
I'm unsure if he used Arabic or Roman notation... Though the dedication page of Summa Mathematica in fact does include recognisable Hindu / Arabic numerals:
https://en.wikipedia.org/wiki/Summa_de_arithmetica#/media/Fi...
Wikipedia's article has several illustrations, but none of the actual bookkeeping notation itself, which is ... a curious oversight. Continued on the double-entry bookkeeping article as well FWIW.
Many people don't realize how fast it is to work on abacus (or soroban in Japanese). Here is an example of what a 7 year old can do
https://www.youtube.com/watch?v=GQtqlB-jXO0
For many years I always did my taxes and other accounting using an abacus because it was sooo much easier than longhand (this was before calculators).
There is the famous Feynman story as well:
https://www.ee.ryerson.ca/~elf/abacus/feynman.html
But you wouldn't impress an ancient Greek very much by taking cube roots quickly because there wouldn't be much call for that.
But there are certainly plenty of mathematical ideas and tools ancient people didn’t know about: they didn’t have a convenient method of manipulating algebraic expressions and equations; they had only the most rudimentary version of differential/integral calculus; they didn’t have group theory, linear algebra, complex analysis, etc.
So could a Roman or Greek. He'd just need a counting board and some stones. He could probably do it faster than you, as well. Anyone using an abacus today could perform almost exactly the same algorithms, with the added convenience of having beads on rods in a frame instead of an unwieldy counting board and stones. The algorithms are the same, and you can also do division and square roots with a counting board or abacus. When done deftly moving stones around on the counting board (and Aristotle makes clear that this goes so quickly that it's possible for the person calculating to cheat an onlooker, like in a con-man's shell game), one might choose to write the result down on a tablet, and he might just finish writing before you do.