Yes, the complex orientation can be factored through these truncations of BP. Either classical methods (the Baas-Sullivan theory of manifolds with singularity - see Baas' "On bordism theory of manifolds with singularities") or more modern methods (see e.g. Strickland's "Products on MU-modules") produce truncated Brown-Peterson $BP\langle n\rangle$ as a tower of "quotients"
$$
MU \to \cdots \to BP \to \cdots \to BP\langle 2\rangle \to \ell \to H\mathbb{Z} \to H\mathbb{Z}/p
$$
and this produces a sequence of compatible complex orientations on these, provided of course that you've produced compatible multiplicative structures on all of the $BP\langle n\rangle$.
The problem doesn't really change if you use $ku$. Also, note that $ku$ and $\ell$ have nicer and more natural multiplicative structures and orientations than any version of $BP\langle n\rangle$ does is known to in general.
