[Release] Update 2.4.0 release logs

[jjyao]

Why are these changes needed?

REGRESSION 11.77%: multi_client_put_gigabytes (THROUGHPUT) regresses from 36.20768872444627 to 31.944310601566972 (11.77%) in 2.4.0/microbenchmark.json
REGRESSION 8.76%: actors_per_second (THROUGHPUT) regresses from 808.4407587905971 to 737.6387503180771 (8.76%) in 2.4.0/benchmarks/many_actors.json
REGRESSION 3.41%: client__get_calls (THROUGHPUT) regresses from 1190.7189254696584 to 1150.072552279968 (3.41%) in 2.4.0/microbenchmark.json
REGRESSION 2.21%: pgs_per_second (THROUGHPUT) regresses from 17.712835160069687 to 17.321330540558634 (2.21%) in 2.4.0/benchmarks/many_pgs.json
REGRESSION 1.34%: single_client_put_gigabytes (THROUGHPUT) regresses from 20.38683304648347 to 20.114238761619227 (1.34%) in 2.4.0/microbenchmark.json
REGRESSION 8983.16%: stage_3_creation_time (LATENCY) regresses from 0.054503440856933594 to 4.9506330490112305 (8983.16%) in 2.4.0/stress_tests/stress_test_many_tasks.json
REGRESSION 996.59%: dashboard_p95_latency_ms (LATENCY) regresses from 12.292 to 134.793 (996.59%) in 2.4.0/benchmarks/many_pgs.json
REGRESSION 358.31%: dashboard_p95_latency_ms (LATENCY) regresses from 592.153 to 2713.875 (358.31%) in 2.4.0/benchmarks/many_tasks.json
REGRESSION 198.77%: dashboard_p99_latency_ms (LATENCY) regresses from 1262.576 to 3772.233 (198.77%) in 2.4.0/benchmarks/many_tasks.json
REGRESSION 23.88%: dashboard_p50_latency_ms (LATENCY) regresses from 3.287 to 4.072 (23.88%) in 2.4.0/benchmarks/many_nodes.json
REGRESSION 20.45%: dashboard_p50_latency_ms (LATENCY) regresses from 4.171 to 5.024 (20.45%) in 2.4.0/benchmarks/many_tasks.json
REGRESSION 10.30%: dashboard_p95_latency_ms (LATENCY) regresses from 1722.469 to 1899.861 (10.30%) in 2.4.0/benchmarks/many_actors.json
REGRESSION 8.31%: stage_0_time (LATENCY) regresses from 11.479018211364746 to 12.432663917541504 (8.31%) in 2.4.0/stress_tests/stress_test_many_tasks.json
REGRESSION 7.71%: dashboard_p99_latency_ms (LATENCY) regresses from 2693.468 to 2901.064 (7.71%) in 2.4.0/benchmarks/many_actors.json
REGRESSION 3.71%: 10000_args_time (LATENCY) regresses from 16.44382056200004 to 17.054044035999993 (3.71%) in 2.4.0/scalability/single_node.json
REGRESSION 3.31%: dashboard_p50_latency_ms (LATENCY) regresses from 3.14 to 3.244 (3.31%) in 2.4.0/benchmarks/many_pgs.json
REGRESSION 2.21%: 3000_returns_time (LATENCY) regresses from 5.873062359000016 to 6.002825282000003 (2.21%) in 2.4.0/scalability/single_node.json
REGRESSION 0.83%: avg_pg_remove_time_ms (LATENCY) regresses from 0.8356835120128541 to 0.8426233363353036 (0.83%) in 2.4.0/stress_tests/stress_test_placement_group.json
REGRESSION 0.27%: dashboard_p99_latency_ms (LATENCY) regresses from 133.659 to 134.023 (0.27%) in 2.4.0/benchmarks/many_nodes.json

Related issue number

Checks

• I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
• I've run scripts/format.sh to lint the changes in this PR.
• I've included any doc changes needed for https://docs.ray.io/en/master/.
  • I've added any new APIs to the API Reference. For example, if I added a
    method in Tune, I've added it in doc/source/tune/api/ under the
    corresponding .rst file.
• I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
• Testing Strategy
  • Unit tests
  • Release tests
  • This PR is not tested :(

[jjyao]

Previous PR: #33905

[rkooo567]

LGTM.