This is exactly what happens, and I notice it myself as I work on my games. On the Simulator, small sprites look quite large because the pixel density of a computer screen is way different than the Playdate. IMO most Playdate games need to be 1.5x to 2x larger in terms of sprite size.
16x16 tiles are too small. 32x32 pixels is just right.
I was doing some tests about this over the weekend with some of Kenney sprites, which are 16x16. Here’s 16x16 tiles on Playdate in the Simulator:
This looks reasonable and legible, right? But it’s way, way too tiny on the device! Nearly incomprehensible. So I tried it out at 2x scale, making the sprites 32x32 in terms of render size:
It almost looks comically large in the Simulator, but it’s actually the perfect size when playing on the device. If it don’t feel too big on the Simulator, then it’s too small! That my motto.
I suspect 24x24 sprites might also be fine and maybe a happy medium. 
I haven’t tried yet though. Game Boy games look big and chunky nowaday for a reason! It was a tiny screen.
This article was written before Playdate came out but is fantastic about comparable sprite size and scaling:
I’ve begun working on my second game for Playdate: Dungeon Quest. I’m giving myself 15 days to try to answer the question: if you remove map exploration from a JRPG and distill it down into combat and some randomized events, is it still fun?
I’ve got the core systems working, like combat, leveling up, strength, etc.:
You’ll descend 8 floors to get to the final boss. Each floors’ events are procedurally generated, so replaying it should be interesting. I’m not sure if it’ll be fun yet though, it still feels like I’m doing plumbing. But I’ve got a handful of ideas for making it interesting. Like a Scavenger who acts as a shopkeeper. And fallen knights, who you can pick up their equipment.
I’ll try to add sprites before my time is up. This is more of a prototype and experiment than a finished game.
I’m trying something new, what I’m calling Game Dev Jump-Start, where I try to finish a bunch of small games with fixed timelines. 2 games a month in March, 2 games a month in April, 1 game in May, 1 game throughout June, July, August. I want to increase complexity but not take on a really long project. And I am still learning, as always. I’m going to focus this exercise on Playdate, so I’ll have more Playdate games to share soon-ish! They’ll all be open source. 
If you want to test out Dungeon Quest or anything else, DM me and I’ll send you builds!
