I also prefer F#'s more general approach, which allows all kinds of extension members (static or instance, methods, properties, and events). But members of the C# team have repeatedly explained that they were the "long pole" during the LINQ work, so I think it's perfectly understandable that they took a more limited approach.
That explanation works for C# 3, but there have been two major releases since, which added relatively little to the language. They could have rounded out the LINQ-only stuff by now.
C# 4.0 brought support for Dynamic, a pretty damn big change. And of course optional and default parameters, another huge boon.
Then C# 5.0 brought Async, a huge feature.
In terms of making the language more powerful I would call those on par with extending out extension methods.
Re:C# Extension Methods Demystified
by magic_haze on Thursday March 28, 2013 @09:59PM
(#5458851)
And I think most of the compiler team has been busy with roslyn, which sounds like a pretty big rewrite.
This is also another area where C# took a narrow view of language features and implemented the bare minimum needed to get LINQ working. Why not take a general view of type extensions, and support type extensions as a proper feature versus a hack for LINQ? That is, support all members, including static members, for type extensions.