I'm trying to understand the result given in the first box at slide 45 of this talk. Specifically:

1. What is the source cited? I have not been able to find any article by Keisler, Chudnovsky and/or Shelah corresponding to the situation.

2. Is this an alternate proof of the classical Morley's theorem (using $\mathcal{L}_{\omega_1\omega}$ as a tool, I guess like how you can use cut-elimination in $\omega$-logic to prove consistency of ordinary first-order PA) or an approach to proving the $\mathcal{L}_{\omega_1\omega}$ version of Morley's theorem?

Thanks.