docs: refresh benchmark comparison with increased message counts

Increase message counts across all 14 benchmark test files to reduce
run-to-run variance (e.g. PubSub 16B: 10K→50K, FanOut: 10K→15K,
SinglePub: 100K→500K, JS tests: 5K→25K). Rewrite benchmarks_comparison.md
with fresh numbers from two-batch runs. Key changes: multi 4x4 reached
parity (1.01x), fan-out improved to 0.84x, TLS pub/sub shows 4.70x .NET
advantage, previous small-count anomalies corrected.

tests/NATS.Server.Benchmark.Tests/CorePubSub/PubSubOneToOneTests.cs

@@ -13,7 +13,7 @@ public class PubSubOneToOneTests(CoreServerPairFixture fixture, ITestOutputHelpe
     public async Task PubSub1To1_16B()
     {
         const int payloadSize = 16;
-        const int messageCount = 10_000;
+        const int messageCount = 50_000;
 
         var dotnetResult = await RunPubSub("PubSub 1:1 (16B)", "DotNet", payloadSize, messageCount, fixture.CreateDotNetClient);
 
@@ -33,7 +33,7 @@ public class PubSubOneToOneTests(CoreServerPairFixture fixture, ITestOutputHelpe
     public async Task PubSub1To1_16KB()
     {
         const int payloadSize = 16 * 1024;
-        const int messageCount = 1_000;
+        const int messageCount = 5_000;
 
         var dotnetResult = await RunPubSub("PubSub 1:1 (16KB)", "DotNet", payloadSize, messageCount, fixture.CreateDotNetClient);
 
@@ -87,7 +87,7 @@ public class PubSubOneToOneTests(CoreServerPairFixture fixture, ITestOutputHelpe
         await pubClient.PingAsync(); // Flush all pending writes
 
         // Wait for all messages received
-        using var cts = new CancellationTokenSource(TimeSpan.FromSeconds(60));
+        using var cts = new CancellationTokenSource(TimeSpan.FromSeconds(120));
         await tcs.Task.WaitAsync(cts.Token);
         sw.Stop();