Looking at the actual cost of formation of fossil fuels makes sense from a total accounting standpoint. That's a depletion of an existing capital stock -- a bank account, if you will. Might want to talk to the good citizens of Nauru about their experience.
And actually, tracing back resources to their antecedents is useful -- there are compounds which formed through biological activity (fossil fuels, limestone, most iron ores, a particularly interesting study), and elements which formed, variously, through stellar fusion (most helium in the universe, though not on Earth, for various reasons, also C, N, O, F, and a few others), supernovae, neutron star collisions, including gold and platinum-group metals (the alchemists were really out of their league...), and, if you want the antecedent of hydrogen itself, yes, the Big Bang.
There are several factors to consider, and source resource costing (solar emergy cost, or exergic potential cost) are among them:
1. Solar (or other source) emergic cost. What was the initial energy influx basis for the resource creation. For fossil fuels, Jeffrey S. Duke's 2003 paper, "Burning Buried Sunshine", provides an excellent breakdown. Humans are burning fossil fuels at the rate of about 5 million years of accumulation per year of present consumption. Given ~200-300 million years of accumulation, and at best a partial recovery rate (not all resource can be feasibly extracted), that's a quantifiably finite period.
2. Production / renewal rate. Another response on this thread looks at water. If you live in a region where rainfall levels are high, say, Seattle, with over 1,000 mm/year, you're starting with a basis of 10 million litre/hectare of water. In Las Vegas, with 100 mm/yr, you're down to 1 million litre/hectare, and have evaporation and soil absorption to deal with as well.
If you're a farmer in western Nebraska, you might want to consider that the water you're pulling from your well represents a few thousand years of accumulation per year of use. That's not going to be particularly sustainable. And figuring your costs only on the drilling and pumping costs, rather than a capital depletion allowance for the water itself, is significant.
While the work which went into creating that resource wasn't contributed by you, it's work that isn't being performed today, at rates equivalent to present usage.
An alternate formulation which might make sense (and this is similar to the description Hotelling used in his original paper, though I came up with it independently and realised that later): say you've discovered a sudden and unexpected inheritance. Not just a rich uncle, but a rich family line has died and left you an accumulated inheritance worth, well, a lot of money.
There's a restriction. You can only withdraw as much of this as you can comfortably carry at a time, it takes some rooting around to actually provide you with the cash, and you've got to take a taxi across town to the bank in order to withdraw your funds.
Does it make sense to account for your cost of withdrawal as only the cab-fare and time costs you incur directly, or to figure in the depletion value of the account itself. I'm pretty sure a bookkeeper or accountant would want to include the latter.
Your next question concerns the size of the inheritance. If it's $10,000, you might spend much of it within a few months, if you were living on it exclusively. At $100,000, a year or two, at $1 million, you could live comfortably for some years, at $1 billion, assuming you were merely spending it down, you could live out your lifetime. The size of the account matters.
(I'm ignoring pollution and other secondary effects here.)
You might even cut others in on the deal if the total value were, say, a few trillions of dollars. Which might make it run down faster.
I'm looking up total resources (that's the total material amount, recoverable or not) of coal and oil. A 1975 USGS estimate was 14.5 trillion tons of coal, and a GeoScienceWorld estimate give 3 trillion barrels of petroleum (http://geoscienceworld.org/content/global-resource-estimates...), though that I believe excludes previous consumption, which I'll assume as that again, so a 6 trillion barrel total original endowment.
If you were to consider the equivalent energy content from what humans previously relied on for biofuel, namely wood, that's ... a lot of wood. A ton of coal is roughly the same energy as a cord of wood -- about 30 million BTU per cord of oak, 14.5 trillion tons is about 13 trillion cords. And 6 trillion barrels of oil is about 7 trillion cords of wood.
Total US annual wood production is slightly less than 20 million cord / year (http://www.fs.fed.us/ne/newtown_square/publications/resource...). I'll assume global wood production is 5x that, for a nice round 100 million cord/year.
We had a 20 trillion cord total fossil fuel resource.
Swapping out that fossil fuel resource for wood production leaves us looking at 200,000 years of equivalent global timber harvest.
Again: we've consumed roughly half that -- 100,000 years of tree growth -- in just over 200 years (and most of that within the past 50). Mind that here we're not talking about input energy (emergy) but the total available energy equivalent (exergy) from fossil fuels. The input would be hundreds to millions of times as much.
(This also has a great deal to inform the question of offsetting carbon output through forestry -- we'd need phenomenally rapid plant growth, and no liberating of that carbon.)
As to your first point -- you're actually correct, we don't necessarily need to figure our cost on the basis of what it took to produce the resource we're using. We can consider the best available alternative fuels, and their cost and flux characteristics. Still doesn't look particularly promising.
Another interesting, though to me fascinating study is to look at the history of extractive resource pricing and price management over time, particularly over the past 150 years or so (the fossil fuel era). Early oil extraction especially was characterised by massive overdrilling -- the "derrick forests" of Titusvill, Oil Creek, East Texas, Los Angeles, and Kern County, all speak to that. Rising alongside this were efforts to create coordinated systems for managing that activity: John D. Rockefeller's Standard Oil, the 1931 initiation of extraction quotas managed via certificates of clearance through the highly inaccurately named Texas Railroad Commission and US Department of Interior (a fascinating history, see chapter 13 of Daniel Yergin's epic on oil history, The Prize, also https://www.tshaonline.org/handbook/online/articles/doe01, https://tshaonline.org/handbook/online/articles/mlc03, and http://www.reuters.com/article/usa-oil-export-controls-kemp-...), as well as national producers and OPEC.
Absent specific limits on rates of oil extraction, in the aftermath of "Poppy" Joiner's "Daisy No. 3" well, oil prices in Texas (and the US) fell from $1/bbl, to $0.13/bbl, and then further to $0.02/bbl, as wildcatters fought to out-extract one another, often on leases too small to individually segregate underground pools. It wasn't until government coercion, at force of arms (Texas and Oklahoma's governors called out their respective national guards, Texas also the Rangers, to sieze wellhead operations), that the overpumping stopped.
Again: without some level of collective, coercive power, the tendency was to simply suck wells as fast as possible, despite falling market prices (lower prices actually promoted ever-more-frantic pumping due to loan obligations), even if that destroyed longer-term productivity and potential of wells. Rockefeller described a similar rationale for his privately organised control system.
Hotelling's 1931 paper alludes to this in its introduction:
"CONTEMPLATION of the world's disappearing supplies of minerals, forests, and other exhaustible assets has led to demands for regulation of their exploitation. The feeling that these products are now too cheap for the good of future generations, that they are being selfishly exploited at too rapid a rate, and that in consequence of their excessive cheapness they are being produced and consumed wastefully has given rise to the conservation movement."
http://www.kleykampintaiwan.com/files/GradEco/hotelling.pdf
When you consider that under-priced energy, again, possibly by a factor of 100x to 1,000,000x, substitutes for labour and depresses the costs of virtually all other production, this becomes more than a passing concern.