The specs and performance don't speak for itself at all - yes it's a fast chip by mobile phone standards, but it's in no way comparable to an x86/x64 processor in terms of computing power.
It also definitely is an ARM - it uses the ARM instruction set, and is based on a licensed ARM design. Yes Apple may have tweaked it and combined it with various other components to form the A11 SoC which is then fabbed by an external company, but at it's heart are ARM based processor cores.
What factual info? One benchmark does not prove performance - when comparing GPUs or CPUs nobody considers a single benchmark to be acceptable, as it is too easy to cherry pick in order to show whatever results favour the author's bias.
If may seem faster for day to day tasks, but that's a result of Apple having full control of both the hardware and software - if you don't have to support 100's of different processors of different ages and capabilities (as Windows or Android both have to) then you can definitely optimise performance for a better user experience.
There is no scientific fact whatsoever. One benchmark proves nothing - come back to me when a mobile phone can mine Monero or other crypto currency at an acceptable rate, perform Raytracing, fold proteins or any other computationally intensive task
The problem is with this debate is - I see numbers, generated on software with known code that proves its calculating through put of data, so the 'one benchmark' is valid.
VS
Your unsound word based reasoning - sure you have some minor/subtle points, because I have presented the evidence (indeed the rest of the world says its comparable performance) and you alone are simply replying 'One benchmark does not prove performance' because that ONLY data we have, disagrees with your personal POV.
Now, I get what your saying, hell I have a masters degree in I.T. - I'm not thick, you make some points, but the degree to which your points are valid is much less than you might be hoping.
Ironically 'Monero or other crypto currency at an acceptable rate, perform Raytracing' is indeed highly hardware specific - and is known to run much slower on intel desktop x64 coding, hence why specific hardware/graphics cards run it better.
Let me remind you, I am ONLY comparing A11 vs Intel mobile - if your running your examples id tell you to run OpenCL hardware!!!
So sir - crypto currency/Raytracing is pointless on intel, nevermind A11.
For running general code - i.e. a processor/CPU, they are benching SIMILARLY - YES(!) its one bench, BUT its running the SAME bench and has PROVEN to be a comparable data set - all other data in geekbench lines up as expected (i.e. running a gulftown intel core vs a sandy bridge vs coffee lake) - all lines up as one would expect in order - the data becomes valid because of 1000's of data points.
Bearing that in mind - A11 is benching ball park i5/i7 mobile chips, faster than gulftown, faster than sandy.
That sir, is why, your comments are just hearsay, and numbers speak for themselves.
