the cost of mining has no effect on the price of coins at this point in time.
Nonsense. Here is as analogy for you. There are millions of 7900 video cards out for sale at retailers, auction sides, and in individual sales, at some average value X. AMD releases a brand new card series, 8900, and prices them extremely aggressively, such that a brand new card is priced below X. Now, since the card is just released, there is limited availability, maybe only 500k cards compared to the millions of 7900 cards already out on the market.
Your argument seems to say that the value of existing cards on the market will not change at all, I call BS. It would be ludicrous to pay X for a 7900 when a faster newer 8900 is cheaper, so the retailers and aftermarket sellers will drop the price of their 7900 cards.
Now, change the situation. The 8900 cards are a colossal failure, they are the exact same performance as the previous generation 7900, but they cost TWICE AS MUCH. Also, nvidia is gone and there is no other source for high end video cards. While the retailers and individual sellers might not raise prices over time, demand will go up as buyers realize that the previous generation is as good as it gets and anything newly made just costs more for the same thing. Over time the cheap older cards are all bought up and the expensive ones remain, but still sell, as long as they are cheaper than the overpriced brand new 8900 cards.
Also, look at history. Blastingcap points out the fall from $30 to 0 to $~2-3 we had the end of summer '11. Look at what happened to mining difficulty! It dropped like a rock,
8/15/2011 141120 1265615 1,805,701 -4.40% -0.30% 14.65 12,925,710 193,885.7 -83,086 $11.3
8/30/2011 143136 1229409 1,777,774 -1.55% -0.11% 14.23 12,725,805 190,887.1 -27,926 $8.78
9/13/2011 145152 1225013 1,755,425 -1.26% -0.09% 14.18 12,565,824 188,487.4 -22,349 $5.84
9/27/2011 147168 1257308 1,689,334 -3.76% -0.26% 14.55 12,092,727 181,390.9 -66,091 $4.92
10/14/2011 149184 1393051 1,468,218 -13.09% -0.81% 16.12 10,509,913 157,648.7 -221,117 $4.01
10/31/2011 151200 1475859 1,203,480 -18.03% -1.06% 17.08 8,614,847 129,222.7 -264,738 $3.19
11/14/2011 153216 1221236 1,192,498 -0.91% -0.06% 14.13 8,536,231 128,043.5 -10,983 $2.51
11/29/2011 155232 1322650 1,090,716 -8.54% -0.56% 15.31 7,807,647 117,114.7 -101,782 $2.79
12/12/2011 157248 1142535 1,155,038 5.90% 0.45% 13.22 8,268,086 124,021.3 64,323 $3.18
12/26/2011 159264 1206050 1,159,929 0.42% 0.03% 13.96 8,303,099 124,546.5 4,891 $4.04
1/8/2012 161280 1122171 1,250,758 7.83% 0.60% 12.99 8,953,273 134,299.1 90,828 $7.05
Despite what blastingcap says about rationality, a lot of miners were rational enough to realize mining at a huge loss was a bad idea. And then when the market recovered, a lot of miners realized that mining was profitable again and it was time to get back into the game. Yes, past performance doesn't guarantee future results, but in the absence of any real hard information this is the best we have to go on.
but that is overshadowed by the 21 million BTC ceiling built into the current BTC network.
They are one and the same. It's not like there is going to be an abrupt end where we go from mining out tons of bitcoins and then suddenly it stops. The step before the bitcoins stop being produced is the step where each reward block offers .00000001 BTC. In other words, based on current valuation, nothing. The step before that is the step where each block rewards .0000002 BTC, nothing. Before that, is .00000004, aka NOTHING. Keep working backwards and eventually you reach blocks where the reward is the equivalent of 1/10 of a cent. The point is, either value goes up massively, or the cost of mining will outstrip the value of the rewards years and years before we reach the last bitcoin.