@halter73?

@JunTaoLuo told me it only made about a 1% improvement in /plaintext benchmark. He's testing again with more headers. If it doesn't make that much of a difference, I think we just close this to keep things simpler.

Here are the results I'm getting from benchmarks:

I tried the plaintext benchmarks with the headers I get from going to github.com:

Cache-Control:no-cache
Content-Encoding:gzip
Content-Security-Policy:default-src *; base-uri 'self'; connect-src 'self' live.github.com wss://live.github.com uploads.github.com status.github.com api.github.com www.google-analytics.com api.braintreegateway.com client-analytics.braintreegateway.com github-cloud.s3.amazonaws.com; font-src assets-cdn.github.com; form-action 'self' github.com gist.github.com; frame-src 'self' render.githubusercontent.com gist.github.com checkout.paypal.com; img-src 'self' data: assets-cdn.github.com identicons.github.com www.google-analytics.com checkout.paypal.com collector.githubapp.com *.githubusercontent.com *.gravatar.com *.wp.com; media-src 'none'; object-src assets-cdn.github.com; script-src assets-cdn.github.com; style-src 'self' 'unsafe-inline' 'unsafe-eval' assets-cdn.github.com
Content-Type:text/plain
Date:Thu, 21 Jan 2016 20:26:04 GMT
Public-Key-Pins:max-age=300; pin-sha256="WoiWRyIOVNa9ihaBciRSC7XHjliYS9VwUGOIud4PB18="; pin-sha256="JbQbUG5JMJUoI6brnx0x3vZF6jilxsapbXGVfjhN8Fg="; includeSubDomains
Server:GitHub.com
Strict-Transport-Security:max-age=31536000; includeSubdomains; preload
Vary:X-PJAX
Vary:Accept-Encoding
X-Content-Type-Options:nosniff
X-Frame-Options:deny
X-GitHub-Request-Id:836BA04A:0179:274B1A9:56A13ED6
X-Request-Id:e52c4a78ec4cfadb14262168197dd21f
X-Runtime:0.011955
X-Served-By:53e13e5d66f560f3e2b04e74a099de0d
X-UA-Compatible:IE=Edge,chrome=1
X-XSS-Protection:1; mode=block

Using pre-encoded headers after about 10 runs:

Median-ish (just eyeballing):
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 6.60ms 12.98ms 229.48ms 96.00%
Req/Sec 16.44k 1.60k 33.29k 77.12%
5268480 requests in 10.09s, 7.51GB read
Requests/sec: 522015.96
Transfer/sec: 761.68MB

Worst:
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 6.48ms 11.11ms 220.26ms 95.86%
Req/Sec 16.15k 1.68k 34.24k 70.72%
5174797 requests in 10.08s, 7.37GB read
Requests/sec: 513315.02

Best:
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 6.82ms 13.18ms 186.46ms 94.99%
Req/Sec 16.92k 1.65k 26.18k 70.80%
5432570 requests in 10.10s, 7.74GB read
Requests/sec: 538024.98
Transfer/sec: 785.04MB

Not pre-encoding after about 10 runs:

Median-ish (just eyeballing):
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 6.43ms 9.50ms 257.14ms 94.64%
Req/Sec 15.54k 1.54k 41.19k 78.03%
4978647 requests in 10.07s, 7.09GB read
Requests/sec: 497220.61
Transfer/sec: 721.13MB

Worst:
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 5.72ms 6.69ms 132.28ms 94.59%
Req/Sec 15.51k 1.85k 55.43k 85.23%
4972067 requests in 10.10s, 7.08GB read
Requests/sec: 492403.45
Transfer/sec: 718.48MB

Best:
Running 10s test @ http://10.0.0.100:5001/plaintext
32 threads and 256 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 6.19ms 9.65ms 189.36ms 95.00%
Req/Sec 15.90k 1.46k 39.37k 75.38%
5094835 requests in 10.08s, 7.26GB read
Requests/sec: 505287.69
Transfer/sec: 737.28MB

This yields about 4-7% improvement when using pre-encoded headers.

After speaking with @halter73 we decided that the minor improvement here isn't valuable ~1% improvement for the plaintext benchmark and ~4-7% when there are many headers.