Thanks for the feedback! We have done all the performance optimization we can in its current manifestation.

We're in the process of refactoring the models that support these visualizations, so we hope to make it more performant in the near future. Hopefully end of this year.