I can't really think of a case where I have made a mistake, but if, for example, I sold something to a friend I would debit (+) their "reimbursement" account with the amount they owe me. If I typoed that to 10x the amount, I would just go and correct the typo, I wouldn't make two new entries to revert then correct the mistake.
Can you share more on the system you've come up with? In what format do you store the data? Do you store files per account, or per day, etc.? Do you have a CLI helper to quickly read, write, edit the files? Maybe you can even share a demo repo?
I believe this same plain text format is used by other tools, which you can find info about here: https://plaintextaccounting.org/ (In particular a lot of people seem to use hledger and beancount)
The ledger is written using a text editor. The purpose of the software is to add everything up, calculate the balances and make sure everything balances. I keep all of my 12 years of accounting in one file and haven't noticed any slowdown. But a real business would surely have many more accounts and may want to split files by financial year or something.
I use helper scripts to convert the data from my bank CSV downloads into ledger format. It uses machine learning to associate payees to accounts (e.g. "Tesco" gets filed to the account "Expenses:Groceries"). I haven't maintained the ML part although it works for me most of the time. In case it's useful, the code is here: https://github.com/georgek/accounts/
I don't bother keeping my ledger immutable, though. The point about immutability is that whatever happened is immutable (because it's in the past; it already happened!) and the ledger should just reflect that. So if I somehow made a mistake in my ledger I just correct it. I keep my ledger in git so that does record changes to the ledger itself, but I rarely, if ever, need to access these.