Like always in 3D printing you need to understand if something is worth printing. There are enough toys that work as 3D print, and enough stuff that either will not survive the load, or not be played with (like those figurines). For those categories (especially the figurines) commercial ones might have the same fate, though, so just printing one to shut the child up may have smaller footprint/costs.
A small list of toys that work very well 3D printed:
This propeller pull toy last longer than the commercial ones with pull string.
This stomp rocket also works great, and if the kids listen to instructions will last ages.
This kind of logic game has similar durability than commercial ones.
This kind of balloon toy also is pretty nice - we used to build those from wood when I was small, but 3D-printing here offers quite a bit more options for experimentation together with the kids. (The author has different models in his profile)
From that there's pretty much a direct line to 3D printed RC models, where the main problem is that many are in the classic model builder mindset where you have to live with the parts you can buy, and due to that end up with a BOM containing dozens of different screw types. This one is an easy to build example not making that mistake, and there are some others as well.