That was the original question, but i am still pondering how "similar" compounds can be, before they can crystallize together.
They need not be similar at all.
For example if a mixture of DMA/DOB HCl is crystallized. Is it possible to make a poly-crystal which consists of both DMA and DOB or will only one of the two things form?
A mix can be obtained, but need not to. The likelyhood of impure crystals (a mixture) forming is proportional to the amounts of the stuff you have in solution to begin with. If you start out with a 1:1 mixture of each, then a mixture is VERY likely to result from crystallization, but if it is a 1:100 ratio, then the resulting crystals are very likely going to be pure (when crystallizing slowly and not crashing out everything). But you can never know for sure until you have tried it.
Where does the line go, when speaking of crystallizing? Can one for instance crash/grow crystals of all the different PEA's in one big batch or what are the limitations?
There is no set line, it is just more difficult to crystallize something pure if you have many contaminants and/or contaminants present in large amounts. But if you repeat a recrystallization over and over, your sample will become purer and purer (in most cases), but you will at the same time experience loss of some for each operation.