It depends what you want to use it for.
Obviously it's good for visualising data in 3d, creating 3D objects etc. and it's not that hard to learn, it's fairly simple if you've programmed before and has a lot of built in functions that make it easier.
The sort of things i've used it for is rendering data from medical xray and cat scans to create a 3D representation of the scanned object which you can then rotate, zoom etc. Also used its BMP scanner which reads in data (we were looking for intensity values) and then renders the BMP with raised high intensity sections and dropped low intensity sections. Using this BMP method with a photo of a human face and implementing the red/blue 3D effect (like what you'd use with red/blue 3D glasses) you can create a cool 3D image of the human face that looks awesome with the 3D glasses 8)
I suppose if you're expecting to create huge, detailed 3D landscapes or something then you'll be dissappointed.
