1) It will take time for Bitfury to deploy that much new equipment. ESPECIALLY with the current yield issues on 14/16nm process from ALL manufacturers on that node.
2) Bitfury has quite a bit of competition already deploying more and more gear
3) Other folks have already announced tape-out or are working on 14/16nm chips
Note that KnC announced tapeout on their 16nm "Solar" chips almost a YEAR ago and announced deployment of those chips a while back, yet their hashrate has NOT exploded since then.
Will BitFury hit 1000 PH? Quite possible.
Will BitFury hit 1000 PH before halfing? Almost IMpossible.
Will BitFury hit 1000 PH by the end of NEXT year? Unlikely.
Will the total hashrate of the entire network hit 2000 PH before Bitfury gets 1000 PH of that deployed? Very very likely.
Keep in mind that there are still SEVERAL folks making chips/gear, BitMain is just the only one CURRENTLY selling to the public (Innosilicon and Lktec have already made announcements, there is that one thread about a rumoured Avalon next-gen machine, Spondoolies has already announced next-gen, plus the other folks that don't sell to the public....)
I think they are just trying to intimidate competitors. There is tons of rumors and promises of projects with no date set. I do not know when the data center will be completed, but its not like they will be able to load up 1000PH of hardware anytime soon.
It is possible however that they achieve it within a year or two, and with their 0.06 chips eventually rolling out, their data center could last quite a while and will then only need to be slowly upgraded, while they phase our their 28nm stuff.
With a project this big, they will probably rent Cloud hash or such. Probably take a page from Bitmain's tricks.
By the time they get their center online, that halving will do interesting things to the market.
But Bitfury will continue doing what they do best, make money