After writing the below, I noticed that you might be asking something else, but anyway, here you go:
The answer depends strongly on whether it is for peer-reviewed, open publications about research, or exploratory development within and secret to a company.
If it is for peer-reviewed, open publications, your question is better approached backwards:
- what is the problem space to study?
- which robot platforms allow me to practically explore it,
- and conduct experiments that are accessible to others?
Robotics research can be categorized into different topic areas (and communities) and various fundamental problems studied in them, respectively. That Reddit user appears to be studying mobile manipulation, so in that case, her needs would include (obviously) an arm and a ROS stack because (less obviously) ROS is prevalent among research papers in that community, e.g., using MoveIt! (https://moveit.ros.org/).
So, to think about Misty for research, think of fundamental problems that are potentially relevant for Misty. For example, in HRI, there is a problem with crowd navigation. Also, there is the problem of navigation where a robot asks humans for help, like “how can I get to Ben’s room?” or “can you open the door for me?”.
The next aspect to consider is ease of use in experiments. Many labs have external motion tracking systems. So, how can Misty be tracked by these? Three common choices are
- planar tracking via 1 or more cameras on the ceiling and patterns on the top of the robot’s head,
- IR LEDs mounted on the robot, and passive cameras mounted around the room,
- small reflective balls mounted on the robot, and cameras with bright IR lights shining into the tracking volume.