This manuscript is an excellent contribution to the DA and geoscience communities.
As all reviewers think highly of this manuscript, I recommend that the manuscript needs significant revision to address reviewer 3's comments and suggestions.
In particular, I agree with reviewer 3's comments that the authors should add a new experiment with nonlinear observation operator to emphasize the robustness of the package in dealing with nonlinear dynamics, which geophysical models commonly deal with.
Also, the Lorenz model is a very comment simple model for evaluating innovative DA algorithms. Therefore, the authors should follow the setup of the popular references and provide the analysis RMSE to give some basic idea about the performance of the new ParticleDA.jl v.1.0 package. Currently, the choice of the observation error variance is very small.

I agree with the reviewer's comments. Although my suggestion for this manuscript is a minor revision, please prepare clear explanations to address the reviewer's comments and revise the manuscript accordingly.

Although I've suggested accepting the manuscript for publication. Some minor comments from the reviewer are provided for the authors' reference.

In my opinion, they have addressed my comments 2-6 well.
For comment 1, I agree with the authors' response, but originally I was hoping they could perform an experiment as indicated by what they mentioned (i.e., use Equation 8 as the proposal density, and use step 5 in algorithm 1 to evaluate the weight).
I still don't see why this is not doable, although doing so is not optimal. It's interesting to see what will happen especially comparing to an ensemble Kalman Filter. That being said, I don't insist they conduct the experiment mentioned above, considering there is already a lot of work done in this paper, and the main objective of the paper is to demonstrate the capability of the software.