SPIKE: Improve TreeView performance

[silvalaura]

This ticket is to investigate ways to improve the performance of the TreeView component. We want to provide adopters a TreeView that with hundreds of nodes, can still be performant. We also need to keep in mind that a multiselect TreeView will especially be slow right now, and we need to consider that use case as well.

We had previously chatted with a product team who wrote up this Wiki and JIRA ticket, but we never followed up with them. It could be worth chatting with them to see if they made any progress that we could work off of.

Some ideas:

• Clean up the code, remove unnecessary functions
• Does virtualization make sense?
• Can we use memoization to prevent unnecessary rerenders?
• ???

Acceptance Criteria:

• Wiki write up on proposal for improvements
• Tickets are created to do this in a phased approach
• Meeting with the team to talk through findings