Tim Anderson has a very interesting blog post with a summary of the history of Delphi for .NET, from Delphi 8 to the coming Delphi Prism. He described the U-turn from "we do the IDE" to "let's use Visual Studio", and the second U-turn from "VCL.NET only" to "use WinForms". But overall, despite errors in the past, he praises Prism and its Mono support. I think that after some mistakes there is now a viable strategy, and that this is not a resource drain for the Delphi R&D team. So it seems a win on both counts. However, we still have to see the product to be able to judge. And to figure out if this product makes sense to the developers, and how many people will actually buy it.
The problems with past version was, as Anderson puts it, "Borland released Delphi 8... Nobody wanted it." Now he specifically points out Cocoa support in Visual Studio as something specific that Microsoft won't probably support, at least not too openly. But is Mono so attractive that supporting it will get developers to Prism? This is not easy to tell...