Hi Tong,
I meant public performance data from regular benchmarkings of public Go benchmark sets, say the 'go1' set.
It could serve multiple purposes, as a source of perf. tuning investigation, for cross-checking with internal data, etc.
I'm not proposing a 'cover-it-all' one but looking for data that could provide supplementary information.
Thanks.