I proposed a variation on that with a SD using the same security (core processing) as the coin. The SD is like a note holding a value which is always less than 1 coins. For any payment the launcher supplies coins and these notes and the core returns one note with the balance that is < 1 coin. Always < 1 coins since the surrendered amount is within a coin of the amount needed. Any return note value of 1 coin is automatically turned back into a coin rather than returning a “note”
When a coin is split into 1 or more notes during a payment, the actual coin is “frozen” by the system. The system can take notes given to it with a value >= one coin and return the coin which was frozen and any left over in a note.
That way the security is the same, the value of any note is relatively small. The network never has more than 2^32 worth of SAFE coin out there.
And it means we can divide the coin as much as we ever want. Maybe initially the division is 1/100ths which would be useful at the moment when coin is worth 10 cents. But tomorrow (with a system upgrade) the note could be 1/1000000ths. Any note at 1/100ths remain the same value because well its the same. just less zeros until it is used then it gets the extra resolution. And we have literally tons of bits to keep dividing. I suggest using decimal divisions since we as humans understand it better, and there is no issue with saving bits because an SD is 100KB, Plenty of room to store the fraction in fixed integer size. Do not use a floating point format because then we will have ultra small fractions of coins lost
Also with safe coin there is an issue of each coin being a physical SD that has to be transacted in an transaction. That is sending ten coins is 10 times the transactions as sending 1 coin. This is why creating SAFEcents SDs is a problem because then potentially sending coinage could be 100 times more expensive on the systems transaction speed. By using divisible notes the transaction cost is only increased by 1 no matter how many “cents” or micro-cents there are.