Faster planning for use cases that does not require a specific end-effector rotation

[Lexseal]

Some users do not care about the rotation of the end-effector, as long as the arm reaches a certain position. This seems like a good opportunity to optimize the time needed to plan a motion.

One idea is to solve the IK without rotation constraints and then feed all the solutions to the planner asking it to return the first successful plan to any of the targets.

Before implementing the whole thing, it is useful to benchmark how much faster IK is without rotation constraints and how much faster planning is when multiple targets are admissible.