Looking forward there are no huge performance gains
14/16nm full custom should be able to get at least double the efficiency of 28nm full custom.
Possibly more, though it appears that quantum effects are making the gains noticeably smaller than JUST the size of the gates would indicate (in theory, if nothing else caused issues, half the size of the gate features should equal half the voltage and a quarter the power, but quantum effects haven't allowed that for a few generations).
THAT is likely to be the last "big gain" for a few years though, as that will bring Cryptocoin mining chips up to the "state of the art".
I suspect Bitmain will skip anyting other than "full custom" on their announced "working on" next generation chip.
I also suspect that chip won't show up before the halfing next year, and might not arrive before end of the year timeframe in 2016.
Look at the CPU stats for your CPU. Look at the GPU stats for your GPU. Consider what "State of the Art" means. If a company can make a 14 or 16 nm chip but the production costs are orders of magnitude higher; this is not a desirable option unless it is cost effective.
When Bitcoin Application Specific Integrated Circuits first began to appear there where many chip manufacturers. This was due to the huge profit margin to be had. That is long gone; we are in the plateau. I do think network difficulty will increase with each new generation of miners; but not like it did in 2013 and 2014. Probably somewhere around 100 PH per generation; with old miner deactivation trailing that increase.
The normal profit model for Bitcoin ASIC has been dependent on the ability to use older tech to make comparatively cheap chips for their mining value. This was possible because of both a high BTC exchange rate and the disparity between chips a company can cost effectively make and their relative efficiency to those already in existence. This disparity is what enabled the violation of Moore's law. The closing of this gap occurred in 2014. The relative depression of BTC exchange rate in 2015 is why you are only now seeing new 28 nm chips being sold. I don't believe Bitmain can double efficiency again with a 28 nm chip ( hats off to them if they can ). Unless the BTC exchange rate goes up significantly I don't believe it is cost effective to produce a 14 or 16 nm asic.
KnC and Bitfury may roi with their own because they have been mining with them longer; but maybe not. I think both of them are more interested in becoming the visa and mastercard of the future via new sidechains and XT add ins. Mining pools will be not only the payment processors of the future; but "processors" in markets that do not currently exist.