I want to add two thoughts to that.
Perspective is important. Something like Diablo 1 & 2 is pixel sprites. No matter how they were created. But the perspective is isometric, making it 3d, while debatedly it is actually 2d, since you could use a top down view like old Moria games. Diablo 3 did not change this perspective thing very much, but the sprites were now rendered in a 3d engine. Starcraft 1 even had height differences and flying units, but all ultimately pixel sprites and iso view.
So 3d vs 2d is related to lots of aspects. Including the engine, the assets, perspective, user interface, level design and real time or not also is a factor of course. There are all sorts of combinations. Like a 2d platformer with 3d graphics for example.
And the other thing is, that the engine and presentation of a game is not important all that much. Bells and whistles can of course help. And are expected in some instances. But ultimately, the game has to be fun enough for people to want to play it. Lacking an advertisement budget and marketing staff, indie devs are wise to do what you described. Finding out, if what they want to realize will garner an audience. Bluntly speaking , there are successful games that are text based. So having nice graphics are obviously not mandatory for success.
But if you do make a 3d game, or any other kind of game of a certain category, you will be measured by the competition in that category. There is less AAA competition in 2d games. So like that saying about not needing to outrun the bear, you only need to be better than the other hobby 2d game makers and small indie developers. And this can be done quickly if you have a good story or mechanic or plain luck and a good enough game. There are examples of really simple games that did go "viral".