Others coming to market with turbines of 1MW or greater include Vergnet, the only French manufacturer and Google-funded Makani Power.
Chinese industrial conglomerate KHCK is reputed to be building a 20MW (!) device which will undoubtedly push the limits of materials science, quite possibly, and literally, to the breaking point.
The economics of wind require larger turbines; hence the race to build them bigger and bigger. The cost of a wind turbine fully installed is now estimated at as much as $4 per watt. The cost goes to $5/W for offshore wind turbines. A big part of this cost, apart from the turbine itself, is the logistics of transportation and installation. As the parts of the turbine keep getting larger and heavier, bringing them from the factory to the installation site is becoming exorbitant.
The sites that have good wind are also frequently at higher elevation, e.g. on ridge lines, where road access and flat staging areas are rudimentary or non-existent. The cost of bringing turbine parts and the enormous cranes used to assemble them high in the air to such challenging sites add greatly to the overall cost. Nonetheless, there are some economies of scale that make it more economical to install larger devices. For example, a 500-ton crane can just as well install a 5MW nacelle as a 2.5MW one. Having undertaken the expense of building a road to the site just to accommodate the equipment one may as well bring in the biggest equipment possible.
So, how big can they get? On land, transportation limits will likely be reached soon, necessitating a cap on the maximum size or a more clever way of breaking the manufacturing into smaller pieces for transport, at the cost of more complicated, time-consuming and expensive assembly on-site. (UPDATE: recent travails of a traveling turbine.)
Offshore wind presents a different set of challenges and opportunities, and manufacturers are increasingly looking offshore for cost and many other reasons. These reasons are why I'm excited about the Grays Harbor Ocean Energy project. More on that in a later post.