Sometimes, a desire to be analytical trumps the desire to simply observe the unintuitive nature of economics[1,2].
> ...abandoned sites express the lack of market demand...
Market demand is subject to the effects of price stickiness[3]. What was once an expensive building on an expensive piece of commercial real estate, might now be very cheap (thus earning demand), but the nominal price remains for some time (this is usually because the company (e.g., Walmart) has enough money that they can sit on an unused asset, rather than listing it at market price). IOW, there isn't market demand, because the nomial price hasn't reached the market price.
1. http://searchdatacenter.techtarget.com/news/4500256609/Green...
2. http://www.containerhomeplans.org/2015/04/what-i-wish-id-kno...
The primary market for abandoned Walmart sites is Walmart', a company like Walmart seeking to compete with it directly. Walmart improves its short and long term prospect by preventing Walmart' from acquiring its former sites cheaply. That's the way real-estate works. Control of a site is a localized but complete monopoly.
If no one is taking over, I'd say it's definitely worth the public resources at least to tear it down.
For most communities, an empty suburban building is not a priority issue.
An empty Walmart could house lots of server racks. The lack of power and bandwidth make the site unsuitable for that use. The lack of cooling equipment and electrical distribution equipment make the building unsuitable for that use. The sea of asphalt and road frontage that comes with it not only isn't worth paying for, it adds to the operating cost of the facility.
These are custom tailored buildings that have reached the end of their viable economic life located on sites where the land was cheap. Their economics was based on operating at a scale that worked with a geographically large but low density catchment.
The abandoned sites express the lack of market demand. In suburbia, there is a cap on demand for 100 acre retail sites. There is a political cost to down zoning high revenue parcels to lower revenue uses and there isn't much demand for those sites anyway: i.e. an industrial or residential use tends not to benefit from high visibility frontage...for an apartment complex it'd often be a detriment.
Finally, the cost of converting a Walmart to a library or workforce housing or a warehouse or whatever is unlikely to vary much from building new. Construction required for conversion from one occupancy to another tends to be expensive. The building shell and asphalt tend to be significantly less costly than furnishings, equipment, and finishes.
Sure there are unicorns where an abandoned retail mecca becomes a vibrant megachurch. But big boxes sit empty because there isn't market demand for another 200,000 square feet of space in that location. Repurposing Walmarts is like housing people in shipping containers. A desire to be clever trumps rational analysis.