As my experience with CFD grows, I start to gain the feeling that turbomacinery is kind of special in CFD: there are special tools tailor for it. For example, ANSYS has TurboGrid or Vista TF for turbomachinery design. And OpenFOAM has special working group for turbomachinery. I wish to know what makes turbomacinery so 'stand out' in CFD technology, I am sure there is something special about it, otherwise we would just use general solvers. So, can somebody shed some light on this mystery (to me)?
1) Turbomachinery is ubiquitous, used for the vast majority of power generation in the world. Therefor, there is a huge amount of money and vested interest in it for practical reasons.
2) Turbomachines are inherently extremely hard to do any kind of detailed experiments on. Visualizing the flowfield or getting any detailed force measurements of an operating turbomachine in order to improve the design is impossible-to-difficult for experiments. This leaves you with CFD as the primary practical tool.
To add to Aurelius' answer, turbomachinery also requires high-resolution rotating grids with somewhat complicated geometries. This is a very hard problem for mesh generation and solvers both.