Spike: Investigate value of Sentry vs performance implications

As part of the FY21 Q3 focus to optimize performance, we want to explore any implications on performance that Sentry may have on the frontend.

The Plan

• Temporarily remove Sentry (!44643 (merged))
• Introduce feature flag to be able to split test performance on the Sitespeed dashboards
• Understand the impact on Webpack bundle and Frontend performance (specifically Large Contentful Paint (LCP))
• Determine strategy for re-implementing error tracking, e.g., include a Sentry webpack bundle on a feature-flag basis?
• Re-introduce Sentry if deemed necessary, undoing changes in !44643 (commits)