Again, those tests are not representative in any way of real-world sustained performance. Someone saying, 'yeah I ran a couple tests in a row' is interesting, but it's not a test.
For a quick look, you can see this chart from last year's macbook test.
http://www.anandtech.com/show/9136/the-2015-macbook-review/9
that article goes into some of the details of why this is.
This article is much more specific to Core M.
http://www.anandtech.com/show/9117/analyzing-intel-core-m-performance/
with something like this being much more representative of Core M's sustained performance (and note that this is only benched against the 15-watt U processors, not the much more powerful processors of the MBP's.
I'm as vocal a proponent of Core-M and the new Macbook as anyone here - I think it's ideally suited to much of the modern computing environment. But I'm also realistic and believe in the right tool for the job. Core-M is not intended for sustained load.