Why did we decide to have a one-to-one mapping between scan and model?

[bjoernenders]

I forget why we did this. It does not appear to make too much sense for ptychotomography. Was it only for better organization of the parameter tree?

[bjoernenders]

And if we emphasize the difference in scans so much, why do we use the same engine and reconstruct the scans together in lock-step and not independently?