It's to do with the lattice energy of the harmala hydrochlorides, along with what is known as the common ion effect. This is how we explain the observed precipitation. Some of the alkaloids do stay in solution after the Manske step, however.
In a way, there is little point in asking "Wouldn't it be more favorable to have the Cl⁻ ions interacting with water instead of MeCOO⁻?" without doing the relevant calculations as that is not what is observed. Besides, the two oxygens in the acetate anion can both form hydrogen bonds with two water molecules. Compare the enthalpy of hydration of acetate anions with that of chloride anions for further insights on this.
It may seem that the crystallisation of the harmala hydrochlorides is entropically unfavoured, but this is balanced out by the remaining solution being a mixture of chloride and acetate.
Also, a methyl group is far from "chunky". You might as well want to consider how acetic acid is completely miscible with water. Or you could try dissolving harmine hydrochloride in glacial acetic acid to see if that precipitates crystals of harmine acetate. (Actually that's not so realistic, maybe salting out from a solution of harmala hydrochlorides in dilute acetic acid using sodium acetate would be a fairer comparison.)
Quote:what keeps vasicine and vasicinone from precipitating?
Firstly, they are typically present only in trace quantities and secondly they would appear to have chlorides that are more soluble than those of the betacarboline salts.
Basic Chemistry “There is a way of manipulating matter and energy so as to produce what modern scientists call 'a field of force'. The field acts on the observer and puts him in a privileged position vis-à-vis the universe. From this position he has access to the realities which are ordinarily hidden from us by time and space, matter and energy. This is what we call the Great Work."
― Jacques Bergier, quoting Fulcanelli