Is there some kind of special graduated engagement required here or something?
What's wrong with something very basic?
Basically, let it flywheel freely (D2 can be any rectifier) by bypassing the load using the switch.
Starting with a naive, simple approach often works well, then improve and refine it to address shortcomings. The problem is these shortcomings are often use case or application specific, such as the specific motor and mechanical load, and when you co-opt someone else's finalized circuit you don't know what their use case or application is, or how well their approach fits
your needs. So you end up reverse engineering it to find its operational limits, response times, dissipation, etc. Or, their circuit is general purpose enough to handle all imaginable shortcomings, in which case it's going to be ridiculously complicated, over engineered, and takes forever to reverse engineer and understand.