I suppose it was a mistake to discuss engineering concepts with a non-engineer. No matter, I will rephrase the question and answer it myself:
1. Suppose my manager wants me to build generate 20kW worth of electricity to run a customer's small business building.
2. It turns out I found a 20kW generator in storage. Now all I have to do is figure out a way to turn the generator and my manager will be happy.
3. I go online and find a wind generator propeller for a few thousand dollars. The next day, Gene shows up at my company trying to sell a PM machine that has 20kW of shaft output power. My manager asks me if it is worth spending money on the PM machine or if we should just go for the propeller.
4. Now I have to weigh the pros and cons of the two technologies. As far as my manager is concerned, the main issue is cost. I don't know how big or expensive the 20kW PM machine is going to be but I can make some guesses. If Gene's machine gives maybe 1% more energy that it takes in, then the machine will be 'taking in' 2,000 kW of power while it puts out 20kW. Because of the slighty over-unity efficiency of 101%, Gene needs to build a machine 100 times larger than actual power requirement. His machine will have the capability to run 101 buildings, but 100 buildings worth of power is used just to run the darn thing!
5. From a cost point of view, it is clear that the machine is going to be rather expensive to buy. Maybe not everyone appreciates what a 2 million watt generator looks like so here's a picture of one:
http://www.servepath.com/dct/05_generator.htm
At this point, the propeller option is looking pretty good.
6. The next day, my manager comes by my desk and asks for my reccomendation. It is going to be very difficult to convince my manager to buy something that is almost as big as the building it needs to power.
My point is that unless the over-unity efficiency of the PM machine is significantly above unity (which it can't be, it won't even be equal to 1), a practical PM machine is going to be huge, monsterous device. It will also be expensive to build, 100x more expensive than a convential machine of equivalent output. If the efficiency of the machine is greater than 101%, say 110%, then the cost will drop roughly by a factor of 10, but if it is lower, like 100.1%, then the costs go up by another factor of 10. Clearly, Gene needs to build something that puts out far more than 1% of the input power.
Another enemy of the PM machine (and other alternative energy sources) is something called the break-even time. If I pay more for the PM machine but it costs less to operate than an alternative, eventually, the lower operating costs will make up for the higher initial cost. Alternative energy sources like solar arrays also have break-even times compared to commercial utility power, and the time is something like 5 years or so. Now suppose the PM machine costs only 10x as much as another 20kW generator. This puts the break-even point of the PM machine 50 years in the future as compared to commercial electricity. This is going to be very difficult to justify from a business point of view.
PM machines, just like all energy sources, need to be competetive with existing technology, otherwise, no one will use it. No one will buy a patent for one, either, unless a cost-benefit analysis shows that the technology will be profitable. A PM machine might be practical if its efficiency approached ~ 200% or so, but until then, it will be very difficult to convince an engineer that the machines are practical.