Multi-files are perfectly fine and perfectly doable, and it for sure is a bunch better if you already have such a system to have that files rather be SQLite DBs than flat files. I assume you obviously have a central DB/repository for central data? If so, having distributed data in a distributed way is just fine, especially if it works for your use case - but if you realize at some point it is cumbersome, then change it, otherwise keep it. The internet is full of people who read "Well it's wise to do X in stead of Y" and then promptly on every forum screams at the top of their lungs: "NEVER DO Y, EVER! ARE YOU STOOPID???? YOU WILL ALL SURELY DIE IF YOU DO!". I take a special pleasure in ignoring them. Other favorites of mine are people abhorring "premature optimization". I agree that premature optimization is a bad habit, but I seem to disagree very widely with some people on what constitutes "premature". Some people will hide behind this little rule to just NEVER do any optimization, ever. if the optimization opportunity is real, you most definitely SHOULD do it. It's one way to tell if a project is maturing: when the optimizations start getting specific. You know your use-case better than anyone on the interwebs - including us. My advice? Try and write down the advantages and disadvantages for any way to do things, and be honest. If you can't think of any, read up online, as widely as possible. Once you have a good honest list, decide what you can and cannot live with or work around long-term, then choose. (And be prepared to have made a mistake and have to change it in future).