Add error context to crash reporting

[albullington]

For better or worse we haven't gotten the 429 errors in Grafana for build 0.59.16 yet that would confirm that this is working in a production environment, but we can simulate that it's working in by pressing the relevant 429 error button. This is intended to help us understand what a user was doing in the app before they receive a 429 error.

[jtklein]

I was not able to trigger the error in the StartupService.js file within a timebox, but I am also sure that the change there is not a regression.

I am seeing some context in the server logs when triggered from logging.js, e.g.
429 in reactQueryRetry: {"queryKey":"[\"fetchTaxon\",48623]","failureCount":2,"timestamp":"2025-04-10T17:41:27.779Z","errorType":"INatApiTooManyRequestsError","status":429}.
I am not seeing context in the server logs when triggered from index.js
{"error":"Too Many Requests","status":429,"context":null}.

My current summary is that this needs more work to be helpful to find patterns of usage that trigger 429 errors.
But I don't see any harmfull regressions here so I Approve.