Perf Migration: Start writing perf issues to the issue platform

[wedamija]

We'll need to start writing performance issues to the issue platform dataset. When we do this, we'll be dual writing for some period of time. Since both perf issue detection and issue platform create/update the issue and counts, we'll need to make sure that only one of these systems is actively updating counts at any one time, otherwise we'll end up double counting.

It might make sense to split up our concept of writing here: Writing the events to the issue platform dataset and creating/updating issues. We can dual write events, since perf issues currently are writing to transactions, and in issue platform will be writing to a new dataset. And we should have a switch that flips us between creating issues via transaction ingestion vs issue platform. Might make sense to make the switch per org so that we can test it in sentry and make sure the issues still look/render correctly.

We should make sure that we send the issue to the issue platform after we've written the transaction to nodestore.