Well I kind of see mining as farming - please excuse how random it is, I'm from Saskatchewan, Canada - pretty much the farming capital of the world.
Anyway, back in the day you could make a living off of a small amount of land. Now a days it's more and more bigger companies owning more pieces of land, and not many people can survive off of farming small time anymore, not like in the early 1900's.
However people who still own land (it's quite expensive as well, a quarter section goes for around $300k, a quarter section is equal to about 160 acres) seem to be renting it out to the bigger players and striking a deal with them, they get paid a certain % of the total yield of that land.
So, I predict mining will turn into how farming is now, except I'm not quite sure how renting land out to bigger plays will relate to mining, maybe smaller miners will rent their equipment out to bigger players for a certain % back? Who knows.
I don't think this is a good comparison. Maybe when it comes to pools this would work in a way but why would someone who can assemble his own chips and miners rent smaller miners? That wouldn't be too smart. I think it will stay as it is right now: Those who produce miners will have preorders, mine with their assembled miners until the ROI is destroyed or at a small margin and sell those miners off. This is the most profitable way.
It might happen to the cryptocurrencies that those people leading bigger assets rent rigs or hashing powers from others but that is the only possibility i could think of. Not everything occuring in the natural world can be transferred to bitcoins and vice versa. (just my 2 Satoshi)
Think of it this way - the value of the old farming land increases, therefore raising the barrier of entry to newcomers (IE people interested in mining, previously CPU or GPU accessibility was common place, ASIC is thousands bare minimum, barring the USB paperweights). The large entities have capital, or acquire it, or source it from smaller farmers who would traditionally harvest the land. Now the land is too expensive for traditional and small farmers to acquire it or the cost of business has risen to a point where scaling economics make more sense (again, GPU vs ASIC..).
It made sense to me, though not exactly the way he concluded it.
However, I do disagree with one point of analysis. As with the small family farmer, the small miner is slowly edged out of the field. Similarly, I do not see "small miners" renting anything to larger entities, rather if anything, they will find other methods to distribute their hardware to turn profit (like ASICMINER's "franchise", allowing a small farmer to get a piece of the pie if they grow the potato on their land, etc). A bit inverse.