but the best fit image will be assigned automatically
There is something that should be clarified. This automation takes for reference the location of first vertex you enter when you draw the patch.
I think developers should rather offer a few best choices according to:
1. Camera closer to the centre of mass of the patch (no to the first vertex), and closest to the principal point of suitable cameras.
2. Best orientation. best fit plane on patch vertices could be used to sort photos according to angle with respect to the patch.
3. Content awareness. In a more advanced approach, the software should be able to skip photos with pixels that are inconsistent between overlapping photos. That would help to remove moving vehicles, for example. Actually I wonder IF Such smart filter could not be used during orthomosaic creation as well to automatically remove people, cars, birds, branches, etc.
Regards
José MartÃnez
GEOBIT / ACCUPIXEL