Page load time is not accurate when after calling newPageLoad

[wvffle]

So basically newPageLoad changes the idempotency, so the same loadTime is registered twice. That way it either lowers the true average time or bumps it up depending on the initial page load.

The problem is, should you count the page load time for newPageLoad as a SPA/PWA page load time (that would require extending the tracking script a bit) or rather don't count it at all for dynamically loaded pages?

Or maybe split it up to First Meaningful Paint and Page Load Time?

[milesmcc]

We could just not pass it when newPageLoad is called, and handle null values on the backend. This feels like the 'truest' solution.