For some reason, I really worry about AMD Fury product competitiveness.

First, Fury doesnt have HDMI 2.0 support. (No support for Ultra HD resolution?)

Second, there is news that HBM cannot be overclocked, it is locked by AMD, uncertain if it is locked by driver or vbios/vuefi (it is possible that AMD already OC it to the max? Or HBM specific interposer issue?)

Third, a reviewer said fury x gpu core voltage is locked as well?

Fourth, there was rumour about 30K units available for year 2015, but AMD denied this rumour. But 10 fury x review units for entire europe......compounded by recent comment by AMD representative about scaling back marketing effort.
Quoted directly ;
"I’m sorry to say that AMD is generally reducing the press events we host, invitee numbers for any events, sampling numbers, etc. Our decisions are based on a number of factors that I’m unable to go into. However I can only assure you that we will involve you in events and sampling when we feel it is appropriate and we are able. With regards to specific sampling of Fiji, I’m afraid our current allocation means we will not be able to sample you prior to launch. I realize this may not be the response you were hoping for but I can only reiterate that we will keep sharing information and more when we can."
I am starting to see the 30K units rumour might be true.

Fifth, when the first news about rebrand broke out, AMD flat outright denied it and say these gpus are for OEM only. -.- Turns out during AMD 16th June 2015 event, it is for official 300 R6 - R9 series as well. Now I am reading about AMD specifically mentions in its reviewer guide not to classify the rebrand as a 'rebrand', instead highlight the so called better binning process, better vram timing and voltage optimization. An apple still an apple no matter how it is genetically modified, no?

Sixth, the latest stuff on driver optimization about tesselation performance specific to 300 vbios.

'Selected' reviewers? Nah.....I dont really care much about kitguru fiasco.

I concerned about AMD scaling down its marketing effort.
Reducing the press events,
Reducing invitee numbers,
Reducing samples sent out,

Sound to me, AMD has serious financial issue right now.
After all, AMD did fired most of its marketing personnels.

As I suspected, GCN architecture really isn't bandwidth limited in first place. HBM is just being used to lower the power consumption and the power saving is reserved for GPU core instead.