According to Lightface mice with finitely many Woodin cardinals from optimal determinacy hypotheses by Yizheng Zhu, theorem 1.1, over $\mathbf{Σ^1_{2n+1}}$ determinacy, $Δ^1_{2n+2}$ determinacy is equivalent to existence of an inner model with $2n+1$ countable Woodin cardinals (which is a $Σ^1_3$ statement). (Per the theorem, it is also equivalent to existence of (countably iterable) $M_{2n+1}^\#$.)
However, $Δ^1_{2n+2}$ determinacy is a $Σ^1_{2n+3}$ statement, and it seems suspicious that over $\mathbf{Σ^1_{2n+1}}$ determinacy, it would be equivalent to a $Σ^1_3$ statement.
Am I understanding this part of this theorem correctly?
Have there been any technical rebuttals to the claim?
If yes, what is the correct statement?
If no, what is the intuition behind it being equivalent to a $Σ^1_3$ statement?