CSHARP-5880: Fix flaky RetryableReadsProseTests.PoolClearedError_read_retryablity_test

[sanych-sun]

The test is flaky on 4.2 variants, as investigation shows even though that the documentation says blockTimeMS for failPoints was introduced in 4.2.9, it does not look like it really works properly on the 4.2 variants:

Test took only 60ms, even though there is failPoint that should block the connection for entire 1s. I've checked multiple 4.2 variants, the test is always takes less then 100ms to complete (both failed and succeeded), when on another versions it took 1s for the test, because of failPoint blocking the connection.

[Copilot]

Pull request overview

Adjusts feature gating to avoid flaky retryable reads prose tests on MongoDB 4.2 where failPoint connection blocking (blockTimeMS) appears unreliable.

Changes:

• Updates the minimum supported wire version for FailPointsBlockConnection from MongoDB 4.2 to 4.4.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

This change intentionally diverges from the documented introduction of failPoint blockTimeMS in 4.2.x by gating the feature at 4.4. Please add a short comment (or a reference to the related ticket/server behavior) explaining why 4.2 is excluded, so future readers don’t “fix” this back to Server42 and reintroduce flakiness.

Suggested change

	private static readonly Feature __eval = new Feature("Eval", WireVersion.Zero, WireVersion.Server42);
	private static readonly Feature __eval = new Feature("Eval", WireVersion.Zero, WireVersion.Server42);
	// Note: although the server documentation lists failPoint blockTimeMS as available in 4.2.x,
	// the 4.2 implementation caused test flakiness. We intentionally require 4.4+ here.