Advantages of Parallella in Applications

Hello,
I have a Parallella-16 Microserver and I want to use it in some small IoT projects instead of a RPI 3, or UDOO x86. As far as I know, due to its 16 Epiphany cores, the Parallella should outperform the other two boards on tasks related to image processing (Kinect, 2D FFT, Face Detection with OpenCV, and Sobel Kernel) and big data analysis (cryptographical operations, video processing) as a whole. Is this really the case, and is it overall a good idea to use the Parallella as a replacement of the RPI 3 in similar cases?
I have a Parallella-16 Microserver and I want to use it in some small IoT projects instead of a RPI 3, or UDOO x86. As far as I know, due to its 16 Epiphany cores, the Parallella should outperform the other two boards on tasks related to image processing (Kinect, 2D FFT, Face Detection with OpenCV, and Sobel Kernel) and big data analysis (cryptographical operations, video processing) as a whole. Is this really the case, and is it overall a good idea to use the Parallella as a replacement of the RPI 3 in similar cases?