We currently have the "Feed Optimization" setting which lowers the feed when a certain threshold in directional change is met. This is suitable for sharp angle changes perhaps, but rather clunky compared to the grace inherent to adaptive clearing.
A more mathematically sound model would instead calculate the cut's true feed rate based on which side of the cutter is engaging the material. When making any arcing cut, using the center of the tool to determine feed is an inaccurate approximation. The outside of the cutter will make a significantly larger arc (depending on the arc angle), thus is cutting at a significantly higher feed rate. Likewise, the inside edge is cutting far slower. An ideal feature would decrease (and increase, as a separate option) feed proportional to these differences in velocity, thus ensuring similar IPT regardless of whether it is an inside corner or a straight edge.
The phenomenon of central vs edge feedrate is well-documented for thread milling, (See: http://www.mmsonline.com/articles/get-the-feed-rate-right-when-thread-milling ), yet it seems to mostly go unmentioned for standard operations. As adaptive and trochoidal pathing becomes more common and the tools spend more time cutting in arcs, its importance only increases. The article linked mentions that some machines may already have mechanisms to compensate for this but there simply is no way your machine can tell for certain which side of your tool is engaging material - thus it remains a problem that only CAM can solve properly.
