I somewhat agree. But there is always a catch to it, then for example it will be more profitable to own a bunch of low-grade computers, and people will thus collect bigger and bigger farms. Distribution is definitely one of the hardest issues to solve.
What I would like to see is a node-based reward system, like Darkcoin's masternode concept, where people are rewarded for anonymizing and fortifying the network. And it should of course also, like you pointed out, be easy and accessible for everyone.
I'm sorry it wasn't clear what is meant by human mineable.
It means a person performs a task on the computer that cannot be performed by bots. It would be cutting the edge of artificial intelligence to keep the coin a step ahead of automators.
1 person sits at a computer.
It could be a 15 thousand dollar computer in Seattle, a 600 rupee used tablet in Delhi, a smartphone perhaps.
They mine manually for x hours and reliably get about y coins. Some people will get y +10%. Some will get y minus 10%. But they all get roughly the same regardless what they can afford.
Will you and I mine these coins? Of course not. The people will be making like 0.1 usd per day. Grunt work.
Will it be a successful algorithm? It is about 100% sure that the first truly human mineable coin will get more publicity than any altcoin since bitcoin. Equally obvious the algorithm will progress similarly to bitcoin. First local game currencies like huc. Next a coin with more progressive intent but still primitive. Then a coin like primecoin that uses the labor, human in this case, more productively then simply running in circles.
The challenge is to develop the coin in a way that stays ahead of automators and to anticipate the effect the coin might eventually have on local economies, i.e., anticipate problems that might arise.
A person could then ask what step would folllow that in altcoins.