Gold can only be predicted with respect to total store, and potentially chemically recreated one for mass production (bear with me on the Sci-Fi a little, I know it may sound stupid, but there is a point - even if it may be incorrect).
Knowing that gold's future supply is variable (as we we use the rate at which we mine it, and the amount we have mined at this rate to predict further stores, AS WELL as the sci-fi point mentioned above), we have to stipulate that Bitcoin has a different context surrounding its valuation.
We know how many maximum bitcoin there will be (in real terms obviously, not nominal terms), we also know that LESS THAN all bitcoin will be recovered (due to losses of wallets and etc.). This FIXED and SMALLER supply has a much different future discount property than the variable nature of what we might perceive our perpetual gold stores to be in the future (again, if we can manufacture that stuff, the model of comparison breaks down).
At this point, to keep it simple and not foray into the territory of delusion (aka my sci-fi comment), we should just take it that the fair value is unknown and that the gold comparison is too linear.
I will add in my opinion that I believe bitcoin is undervalued for the fact that there has been more activity in the sphere despite the lower price, and that if you aggregate the salaries of the talent pursuing the industry as well as the opportunity costs of the venture capitalists supporting them, then the price of bitcoin is not as high as it should be. Furthermore, there is a higher level of acceptance for bitcoin from big 1% companies (microsoft for example) than there was at the same time last year when the price was higher. Add to this that the current drop in price since December 2013 though large, in relation to prior drops (I remember when it went to $8 like two-three years ago) is actually smaller in magnitude and therefore implies a lesser degree of volatility.
That's my two cents, please give me some activity points LOL.