Benchmarks are not accurate for real use

[Martmists-GH]

Please update benchmarks to test for (and put these separately in README):

• simple text
• wall of text
• simple html
• wall of html
• images of various sizes
• gifs of various sizes
• big files (e.g. zips)
• backend processing (e.g. API)

[tejastank]

WITH SIMPLE TEXT THIS FRAMEWORK IS HIGHLY BETTER BENCHMARK.

YA. ITS POOR DOCUMENT AND CORE TEAM VERY POORLY MAINTAIN AND UPGRADE.

NO SOLID PLAN FOR FUTURE UPDATE OR NEXT NEW CORRECTION.