fix(core,browser): Delete SentryNonRecordingSpan from fetch/xhr map

[JPeer264]

This resolves a leak where SentryNonRecordingSpan are pilled up when tracingSampleRate is set to 0. Theoretically SentryNonRecordingSpan are still treated as spans and added to the spans list, but never removed

By moving shouldCreateSpanResult closer to the actual span logic, this is now resolved.

Closes #19337 (added automatically)

[ovflowd]

TSYM!

[Lms24]

Good catch!

[github-actions]

Codecov Results 📊

---

Generated by Codecov Action

[github-actions]

node-overhead report 🧳

Note: This is a synthetic benchmark with a minimal express app and does not necessarily reflect the real-world performance impact in an application.
⚠️ Warning: Base artifact is not the latest one, because the latest workflow run is not done yet. This may lead to incorrect results. Try to re-run all tests to get up to date results.

Scenario	Requests/s	% of Baseline	Prev. Requests/s	Change %
GET Baseline	8,962	-	8,792	+2%
GET With Sentry	1,539	17%	1,673	-8%
GET With Sentry (error only)	5,883	66%	6,131	-4%
POST Baseline	1,154	-	1,193	-3%
POST With Sentry	559	48%	594	-6%
POST With Sentry (error only)	1,023	89%	1,046	-2%
MYSQL Baseline	3,086	-	3,254	-5%
MYSQL With Sentry	299	10%	462	-35%
MYSQL With Sentry (error only)	2,553	83%	2,628	-3%

View base workflow run