options I see would be using oxygen inhibition layer like carbon..  or multiple colours and an intersection bundle like xolo lithography

I guess it's finally time to print the missing parts and assemble a prototype, see how single layer passes work.

i would start with one or two laserdiodes... image directly onto a ccd camera  with neutral density filter..  that's how I started with my prism scanner. You would have to argue against the prism concept or find an application with best fit. There must be at least one.. a problem You need to solve is getting the correct focus for all your laser diodes. Due to manufacturing tolerances this is hard