Here is my weigh-in on the topic . . .
This has been a very interesting discussion that underscores one very important point: block size is an extraordinarily complicated issue that probably has been influenced by a number of different factors! Which factors get activated (and how and why they do) is going to vary from place to place, making a one-size-fits-all explanation an exercise in futility.
Going back to the original blog entry (the accompanying poster/graphic of block size variations in U.S. cities is fantastic and speaks to my geo-geeky soul!!), there may be some credence to the idea that climate/local environmental conditions (carrying capacity) has something to do with block size in *some* places--but not in the way that the original blog entries argue.
In the American Great Plains (where I grew up), towns were often planned slightly in advance of agricultural settlement--largely as speculative real estate ventures. Some "developers" had little to no experience or understanding of what they were doing and simply threw down a grid (or some other geometric scheme) within the large plot they had purchased within the U.S. rectangular land survey system, hoping that their cartographic expression of established, eastern urbanity would convey the message that "this place is going to be a success!"
Whether or not their schemes worked depended upon three things: 1) emerging rural settlement and agricultural productivity levels in the hinterland immediately surrounding the new town (which was to a large degree dependent upon environmental/climatic conditions), 2) transportation/communication connections between the hinterland/new town as well as the new town/larger market places elsewhere, and 3) affordability/potential value-added derived from owning property within the town. If the hinterland was productive, the connections were good, and the property appropriately priced, then the town/real estate venture stood a chance because there would be demand for the lots. If any of those things didn't fall into place, the town might languish until external conditions changed. Or, it might fail altogether (Mark Twain wrote quite gleefully about some of the places that failed in *Life on the Mississippi*).
The larger point is that other speculator/town builders were really savvy about all of this. They understood the boosterism that needed to surround the launch of a place, just as developers understand the importance of branding their real estate projects today. Building the image of success was critical. Having a town plan that emulated other ALREADY SUCCESSFUL places was often a part of this.
They also understood the importance of attracting merchants--so, as Dennis McClendon pointed out in his post--things like corner (and oversized but cheap) lots become really important toward INCUBATING what ultimately becomes Main Street/the CBD. Plus, they understood the delicate dance between pricing the residential lots low enough to attract the critical mass necessary to get the town off the ground and making enough lots and charging enough for them to cover the initial outlay for the land.
The calculus of how many lots were going to be needed to recoup the original investment and generate some profit in conjunction with the 'civic impression' the developers wanted to make ('high density [small lots] = good impression because we're hustling and bustling' or 'density = bad impression because now we look like those horrible places Back East', 'wide streets = bright future' v. 'wide streets' = too much land removed from potential profit making) might have more to do with the variation in both lot and block size in many places than anything else.
Historians and historical geographers who have written about this in some fashion include Bill Cronon, *Nature's Metropolis*, Tim Mahoney, *River Towns in the Great West*, Richard Francaviglia, *Main Street*, and Hildegard Binder Johnson, *Order Upon the Land*.
I also wrote about it in my work on the model industrial town outside Pittsburgh that the Olmsted firm planned during the 1890s: Vandergrift, PA (*Capital's Utopia*, Johns Hopkins University Press, 2004). In fact, this whole issue of lot and block size became a major bone of contention between the Olmsted firm and their client, the Apollo Iron and Steel Company.
The steel company had initially said that it wanted the Olmsteds to create "something better than the best" existing industrial settlement ANYWHERE, and they wanted a full complement of infrastructure to be a part of that. Given that this was all happening right after the Pullman and Homestead Strikes, the company also wanted to minimize its long-term engagement within a residential rental market. Homeownership for the workers via the provision of mortgages was going to be a part of this plan.
The Olmsted kids (John, Rick [to some degree], and Charles Eliot--FLO Sr was in the sanitarium by this point) responded by creating this beautiful plan with wide streets, a village green, BIG BLOCKS, large lots, ample lobes to the curves in the streets, parklets punctuating major AND minor intersections, etc.
But . . . when the company sat down with engineers about the sewers, watermains, sidewalks, and street lights, found out how much all of that was going to cost, and then did the math based on how much they would have to charge per lot in the Olmsted plan, they realized they were never, ever going to be able to break even--because steel workers would never be able to afford those lots. Thus begins months of testy negotiations between the company and the Olmsteds over how to tighten up the street grid and the lot sizes within the grid without sacrificing the infrastructure and all of the benefits of having an Olmsted-esque plan.
You can see tangible results of this negotiation in the Olmsted's final presentation of the plan. The streets and lots on the left-hand side of the map were surveyed as mapped and were the first section of the town to be built. The right-hand side (the really curvy streets with all of that open parkland) were part of the original Olmsted plan but were never built that way. Notice the huge lot and block size in comparison to what actually was created.

In the end, John Olmsted was really unhappy with how the town turned out--he thought too much of the original vision had been sacrificed to economics. The steel company (and ultimately the residents) were pretty happy, however. The lots sold, costs were recouped, the steel company could handpick who it got who to sell to (non-unionists for its mill!) and the place looked far better than any other steel town in western PA--there were still curvilinear streets, some parklets, a small village green--and all of that infrastructure, including indoor toilets! Plus most of the housing was owned by the workers. Because of all of this, Ida Tarbell (who had been very critical earlier in her muckraking career about capital's treatment of labor) ended up declaring Vandergrift the most important industrial town in America because of the good 'social contract' it represented--even with the small lots and blocks!
Finally: here's one more interesting complication to the block/lot size conundrum. James Vance speculated in his book *The Continuing City*, that lot sizes (and hence, block sizes) in many European cities were influenced by the MODAL SIZE OF TREES in forests surrounding nascent towns in the Middle Ages. When towns were growing 'organically', tree size = timber length = beam length = lot width!