While we made significant improvements to the store cleanup in #187, the performance tests didn't really reflect an improvement in speed. This is because while we create lots of elements, those elements are very bare bones, and don't have state.
We should create a performance test switches between a large number of elements with their own state controls. Something that specifically performs slowly before #187 and quickly after #187
Summary
While we made significant improvements to the store cleanup in #187, the performance tests didn't really reflect an improvement in speed. This is because while we create lots of elements, those elements are very bare bones, and don't have state.
We should create a performance test switches between a large number of elements with their own state controls. Something that specifically performs slowly before #187 and quickly after #187