You could give each note a unique identifier, which would be appended to the beginning of each line within a single text file. Then use filter to pull up the relevant identifier, and discard the identifier from the output. This could be achieved by using filter -k and filtering through an alias which does a tokenize 32 on the input text, then returns $2-.

Then you would just need to make a system to make the notes manageable to the script and be user-friendly. Perhaps requiring them to enter a title on their note(s), which would replace spaces with underscores. Then, the title (with underscores) could be used as the identifier.