I tried to get some "raw numbers" that quantify a bit the balancing situation we are discussing. As an example, I examined the following city (which seemed to me not terribly biased to any direction):
1 +3 Food Resource
1 other resource +1F (perhaps a cow)
8 grassland tiles, half of them next to a river
6 plains tiles, half of them next to a river
Of course I don't claim that this is a completely average city - I just picked it because I happened to have 2-3 such sites in my last game. Then I also used as working hypothesis that this city gets a granary as soon as it gets to population 4. With these conditions, I measured commerce earned for 256 turns (binary arithmetic is a bad addiction) by cottages for a non fin civ, a fin civ, and a "modified fin" civ with 1 commerce for each 3. Since 256 turns might not be enough to get to Printing Press, I considered Villages/Towns without the additions. I also checked the possibility to use only specialists in this city. The results were the following (as of turn 256):
Non fin civ with cottages: Commerce per turn 63, Total Commerce earned 10788.
Fin civ with cottages: Commerce per turn 77, Total Commerce earned 13572. Advantage over non fin civ: 22% in commerce per turn, 26% in overall commerce (total difference 2784 commerce!!!!)
Modified Fin civ with cottages: Commerce per turn 77, Total Commerce earned 13257. Advantage over non fin civ: 22% in commerce per turn, 23% in overall commerce (total difference 2469 commerce!!!!)
Since these numbers just come from one not exceptional city, and without even continuing to count for more turns (I used Epic speed, and Epic has 660 turns totally) it seems to me that the Financial advantage, even with the modification, is too much. Imagine only that this 14 commerce per turn will continue to add for the rest of the game. Maybe pop-rushing in some points would create some difference in the numbers, but the value is just too high.
These numbers also provide a good answer to whether Financial is more powerful than organized: you need only something like 8 such cities to overcome the Org discount for a civ with 28 cities, based on the comparison data from the Apolyton site - not to mention that with the Fin civ "as is" I could get much better results if the plains were water tiles!
Next I tried to count the specialists approach for this city (I had calculated it before, but my numbers were wrong because I hadn't included the granary). The numbers are terrible: 9190 Beakers (assuming Pyramids/Rep), i.e. 1598 less "output" than the cottages (and this from a single city). The only consolation is the 2752 GPP they gother (good for 5 GP). If it was the only city that would be more than enough - but if you try to do the same in many cities, you will not get much more GP.
For a Phi civ you would get 5504 GPP (8 GP from a single city). This is better, but still can't be easily scaled for many cities.
PS. In the calculations I ommitted the central tile of the city, because it doesn't make any difference in the evaluation (it only alters very slightly the percentages, but you can count it also yourself).
EDIT: Corrected the numbers for the specialist approach. I mixed up the XL files and mistakenly took the data from the "non rivers" version. The correct output is 9190 "Beakers+Commerce" for this specific case and 8256 for the "no rivers" case. One obvious observation is that, in the case of a city with no rivers, the difference becomes quite small.
