Developper-created currencies always suffer from a massive problem: they model their currencies from real-world money, then find out that real money does not work well in a small-scale economic simulations. Too simplistic, and stuff doesn't work, too realistic, and the best way to get rich would be to strike rich, then invest and relax (instead of playing).
The interesting part when you explicitely do NOT create a player currency is that you start with barter, and the player economy quickly settles on using a "main good" as a currency proxy. The main good is whatever is simultaneously difficult (but not too much) to get, and useful for almost everyone: i.e. one that has a good balance of faucets and sinks. You don't have to painfully balance your sinks and faucets: the player economy will find out which one works.
You still need to be careful about your goods balance, but you don't need to balance that ONE good. The players will find out which one you balanced well.
I'm going to take The Settlers Online as an example. In TSO, you have a barter market: you can sell anything and ask for anything in exchange. Which means your market sees a lot of barter offers (my marble for your bronze swords). There's a gold coin, and it has a good popularity because people expect it to be "the currency", but the real currency in-game is the Granite bloc. Granite is required in quantities to upgrade to high levels your buildings. Low-level players can get small quantities of it, but have little immediate need for it because they haven't upgraded their economy enough to start needing it, high level players get more, but have lots of buildings that can be improved.
For a time, a new resource looked like it might upset the market (the magic bean), but while it has traction, it lacks the standing power of Granite because only the high level players can get it (you need level 48 to enter hard adventures that reward it), and the uses for it are relatively few. It works due to rarity, but fails as a main currency: most trades are granite-oriented.
Note that you need to design your game to work for that. Your goods must be heavily divisible, or barter does not work well. If you have large blocks (something Crowfall seems to be leaning toward), you have difficulty using it as barter, plus it's very difficult to tax. However, if it's heavily divisible, you lose the caravan aspect: people can stuff their pockets and make multiple low-risk/low-reward trips instead of putting all their baskets into a high-risk/high-reward caravan escort, and most people are risk averse.