fix(ios-tests): reduce testCleanResyncGhosts flakiness

[wmathurin]

Summary

• Parallelize independent netCreate calls — the two contacts are independent; running them with Promise.all saves one full server RTT off the critical path
• Add 1s pause after netDel before cleanResyncGhosts — Salesforce has eventual consistency; the deletion needs a moment to propagate before the ghost-detection query runs (mirrors the existing pattern in testReSync)
• Add .catch() handler — without it, any transient network error causes the test to hang silently for the full 120s XCUITest timeout instead of failing immediately with a useful message
• Fix wrong assertion message — 'Total size should be 1' → 'Total size should be 2'

Context

testCleanResyncGhosts has been the sole flapping test in the iOS suite for a while. It timed out in CI on both the ^18 and ^26 legs of PR #477. The test makes 6+ sequential network calls to a live Salesforce org; on a slow CI runner that easily lands on or past the 120s XCUITest timeout. The parallel creates address the root timing cause; the .catch() ensures that if it does time out in future, the failure is reported immediately with a message rather than as an opaque hang.

Test plan

• Run iosTests locally: xcodebuild test -workspace SalesforceReactTestApp.xcworkspace -scheme SalesforceReactTestApp -destination 'platform=iOS Simulator,name=iPhone 16 Pro'
• Confirm testCleanResyncGhosts passes and completes faster than before
• Confirm all other MobileSync tests still pass

[github-actions]

	Tests	Passed ✅	Skipped	Failed
iOS ^18 Test Results	35 ran	35 ✅

Test	Result
No test annotations available

[github-actions]

	Tests	Passed ✅	Skipped	Failed
iOS ^26 Test Results	35 ran	35 ✅

Test	Result
No test annotations available