The sim2real approach can work pretty well as long as you are very in tune with where your simulator falls short relative to the real world and take steps to circumvent those shortcomings.
We were able to train the robot to climb stairs completely by feel/proprioception without any sort of vision. We trained it in simulation, and then transferred it to the real world without issue.
Here's some work my lab did on sim2real for a roughly human-scale bipedal robot (Cassie): https://www.youtube.com/watch?v=MPhEmC6b6XU
We were able to train the robot to climb stairs completely by feel/proprioception without any sort of vision. We trained it in simulation, and then transferred it to the real world without issue.