Just a quick reply ...
I insist, neither blockchain statistics nor trade volumes give us any useful insight on actual adoption as means of payment. The traffic in the BitPay and Coinbase "wallets", as gathered by that Czech site, would be a more direct evidence (assuming the wallets were correctly identified, which I cannot tell for sure). As I wrote, I have had a look at the Bitpay wallet, and did not see growth in 2014 either. But please check yourself.
Don't know the the Bitpay wallet statistics, but would be interested. Any link for it?
Anyway, in my opinion there are two general problems with your counter-arguments to the adoption evidence.
(1) Presented with several metrics that intuitively seem to be half-decent proxies for "adoption", you find fault with any of them on the basis of what I'd call minor problems. But maybe I get what bothers you...
If it makes you more inclined to see it my way, I will admit this much: I agree. We cannot be absolutely sure adoption is growing based on those statistics. They could be (partially) fake. They could over-represent non economically significant transactions. And so on.
However, they are the best approximation
we have, and by and large, they point to increasing adoption. So while I agree that there remains some epistemic uncertainty perhaps, there is little aleatoric uncertainty that adoption grows given this data.
(2) You don't really follow the Copernican principle applied to time when you only concentrate on a very narrow range in the very recent past where adoption did not soar like it did before, and conclude from it that there seems to be no further adoption - ignoring that overall the adoption metrics are significantly up compared to where they were a year ago, and
vastly up from 2 years ago.
Okay. Here's another one, this time, pure merchant driven transactions:
In May 2013 Bitpay had a volume of 5 million USD per month processing transactions on behalf of their merchants.
(source)A year later, May 2014, it's 1 million USD per day.
(source)Time for the same spiel again? "Maybe it's up from May to May, but how can we be sure that from May to
now adoption continued?!" Yes, we can't be sure. But given the (often incomplete, chunky) data, I consider the hypothesis "growing adoption, though maybe not growing as fast as in the very early phase" more likely than "sudden and complete saturation/end of growing adoption".