I would add all your scatter geometry to a single tyMesher (in "input geometry" mode), then use a Script operator to cast rays from your camera to each desired pixel, checking for hits on the tyMesher.
The easiest way to setup the raycasting would be to create a plane and orient it to the camera...offseting its position so that you can see it in the view and it perfectly fills the screen. Then set the plane segs to the resolution (so it has the same number of verts as the desired number of pixels). From there, use a Script operator to cast rays from the camera position to each vertex, and position a particle at the hit location if a hit is detected on the tyMesher. You can also scale the particles based on camera distance so they're all the same size in screen space.
The easiest way to setup the raycasting would be to create a plane and orient it to the camera...offseting its position so that you can see it in the view and it perfectly fills the screen. Then set the plane segs to the resolution (so it has the same number of verts as the desired number of pixels). From there, use a Script operator to cast rays from the camera position to each vertex, and position a particle at the hit location if a hit is detected on the tyMesher. You can also scale the particles based on camera distance so they're all the same size in screen space.