feat: add benchmark test project for Go vs .NET server comparison

Side-by-side performance benchmarks using NATS.Client.Core against both
servers on ephemeral ports. Includes core pub/sub, request/reply latency,
and JetStream throughput tests with comparison output and
benchmarks_comparison.md results. Also fixes timestamp flakiness in
StoreInterfaceTests by using explicit timestamps.

tests/NATS.Server.Benchmark.Tests/Harness/LatencyTracker.cs

@@ -0,0 +1,51 @@
+using System.Diagnostics;
+
+namespace NATS.Server.Benchmark.Tests.Harness;
+
+/// <summary>
+/// Pre-allocated latency recording buffer with percentile computation.
+/// Records elapsed ticks from Stopwatch for each sample.
+/// </summary>
+public sealed class LatencyTracker
+{
+    private readonly long[] _samples;
+    private int _count;
+    private bool _sorted;
+
+    public LatencyTracker(int capacity)
+    {
+        _samples = new long[capacity];
+    }
+
+    public void Record(long elapsedTicks)
+    {
+        if (_count < _samples.Length)
+        {
+            _samples[_count++] = elapsedTicks;
+            _sorted = false;
+        }
+    }
+
+    public LatencyPercentiles ComputePercentiles()
+    {
+        if (_count == 0)
+            return new LatencyPercentiles(0, 0, 0, 0, 0);
+
+        if (!_sorted)
+        {
+            Array.Sort(_samples, 0, _count);
+            _sorted = true;
+        }
+
+        var ticksPerUs = Stopwatch.Frequency / 1_000_000.0;
+
+        return new LatencyPercentiles(
+            P50Us: _samples[Percentile(50)] / ticksPerUs,
+            P95Us: _samples[Percentile(95)] / ticksPerUs,
+            P99Us: _samples[Percentile(99)] / ticksPerUs,
+            MinUs: _samples[0] / ticksPerUs,
+            MaxUs: _samples[_count - 1] / ticksPerUs);
+    }
+
+    private int Percentile(int p) => Math.Min((int)(_count * (p / 100.0)), _count - 1);
+}