Leverage WebWorker for clustering

[Etienne-M]

Clustering a large amount of probes can take ~1 seconds by itself. All of this time is currently spent on the main thread, which yields a poor user experience.

We could move this computationally intensive logic inside a web worker to free up the main thread.

This issue has already been opened in the past, but it was previously closed: #131

[ykzeng]

Hi @Etienne-M this indeed sounds like a significant amount of refactoring as previously mentioned. Perhaps we can use this issue as a poll for such improvement. If we got a sufficient amount of attention we can def prioritize it. In the meantime, feel free to open a PR if you had any solution as well we can review it.