Motion Tracking: Survey Shots & Known 3D Locations

Blender's motion tracking functionality could really benefit from using survey shots and the ability to attach known 3D locations to markers prior to solve. These functions exist in professional motion tracking software, and Blender is SO close to being able to compete, but just missing these two crucial features. I'll explain what I'm imagining, and then why I believe that's important below.

Survey Shots

Blender could add a survey functionality by allowing you to "share" markers between clips that represent the same 3D location in a scene. The solver would then factor in any "shared" markers from any other clip in your project to aid in the solution. This would be similar to how Blender uses data-blocks and being able to use the same mesh data in different objects. Each marker could have a unique marker data-block, or be linked to another existing marker data-block from another clip. Behind the scenes, the solver would then essentially treat any clip with shared markers as one solve, using all the frames and relevant shared markers from each clip as one (compensating for different aspect ratios/frame dimensions of course).

Why Survey Shots?

Survey shots are used in professional tracking software (PFTrack, Syntheyes, 3DEqualizer) to aid in the solve of challenging shots, especially when there is low/no parallax information, and to align multiple shots in the same scene. Basically additional survey photos and video from various angles of the location can be used to give the software additional parallax that it doesn't have in the primary shot. These survey photos and their markers can be used for any shot of the same location, which helps in aligning multiple camera angles from the same location, or shots that have limited parallax.

Known 3D Locations

Blender could add the ability to input known 3D locations to a marker, either by manually typing in the coordinates, or by attaching a marker to an existing empty in the 3D scene. This is different from generating an empty from a marker AFTER the solve, as this is used for the initial solve itself, to give the solver more information about the 3D scene. It would probably be added to the "Track" tab for a selected marker, with a Target picker option and Location XYZ fields for inputting manually.

Why Known 3D Locations?

Similar to using "shared" markers to aid in the camera solve, giving the solver known 3D locations for a selected marker can also significantly help solve for the camera in challenging shots. This comes in handy when you have photogrammetry / LIDAR data of the location, or have just previously solved cameras in this location and have reconstructed geometry of the scene. You would be able to place the 2D markers exactly where you know they should be in your 3D scene prior to solving. This is useful when you have numerous shots to solve in the same location all at various angles and need them to all be aligned in the same 3D space, or when you have low / no parallax in a shot but could still localize the camera by identifying known 3D locations in the frame. Object tracking would also benefit greatly from using known locations, as you could use locations on the surface of a mesh that represents a real world object to aid in the solve.

I hope I've explained myself properly with both what I'm imagining and why I think it's important. I would love to know if anyone else thinks these features are important and how you might implement them.