I have to say, I do see the OP as raising an interesting and valid point that at some stage, donating cash may be more beneficial than purchasing
[additional] hardware (over and above what you would purchase for everyday use) and paying the ongoing operating costs. Indeed, with the improvements to the energy efficiency of computers when idling and power prices generally increasing faster than inflation, I can only see it becoming more so, to the point where it could be more cost effective to idle your computer and donate than to run it flat out.
I think it has two separate aspects - donations for the running of FAH and donations for the funding of actual folding machines. In the case of the former, I suspect it would depend on whether the scientific goals of F@H were more constrained by 'operating' costs, or by the amount of folding being done by the community.
For the latter, I see it as a straight substitution. I.e. whether the performance of $1000 of folder computer & energy is >=< than $1000 of F@H operated computer & energy. For one thing, donations to F@H are tax deductable, whereas purchasing personal machines are not (there maybe some exceptions around companies etc). I imagine F@H, as part of an organisation that purchases/leases a great deal of computing resources potentially has access to deeper discounts on hardware than regular folders. This could skew the equation towards donating for hardware, at least for computers over and above those used for everyday folding. As computer idle efficiency improved this could be further skewed (if a modern GPU consumes less than a watt while idling, then the operating costs associated with running it at 100% constantly may greatly offset any sunk costs) especially as the cost of new hardware falls (reducing sunk costs) and more portable devices are favoured*.
I suspect where it would balance out is whether they have to purchase enterprise-level/quality hardware, or whether they are able to purchase consumer grade hardware. A further limitation would be the time factor - maintaining racks of user funded folding rigs would require human resources that, when run by the community, are provided by the individual. I am sure there are many other examples of reduced 'efficiency', such as the non-monetary benefit folders get from playing with their fancy computers - as illustrated by
DoctorsSon 
How energy efficiency gains due to process shrinks or transistor design influences the 'cost' effectiveness of whether to idle everyday hardware & donate or fold I am unsure - perhaps if the energy efficiency improves faster than power prices and the energy consumption delta between idle/100%.
I would suggest that if idle & donate becomes more 'cost' effective than donating computer resource time then the validity of the distributed computing model would be up for debate.
The dollars for points is an interesting idea - pool a number of users to purchase a more powerful computer than they would, on their own, purchase and leverage the QRB for higher ppd per user from it... However, the difficulty with such a proposal is that over time the 'subscribed' computer will decrease in cost effectiveness and maintaining such a system would be ongoing recalculation - how would you fund ongoing costs while removing the 'ability' for folders to cut and run once more efficient computers are available?
mdk777 wrote:Any HP accounts finding it still pays to rent rather than own?
Is anyone actually running HPcloud since the Free public beta ended?
*Computer users moving away from $1000 PCs to $600 tablets is a 'risk' to the current growth of the F@H model.
/End my lengthy opinion piece
