Ben, 
 I am afraid that neither I nor Xiangrui Li (who develops dicm2nii) have access to Philips hardware. Further, the Philips PAR/REC is proprietary and we do not have a formal documentation (and the implementation has changed a bit over the years). The main issue seems to be that your PAR file describes the "image_type_mr" as being "4" and I have no idea what that means. Most MRI scans only store the magnitude image (type 0). However, field maps often store the the magnitude and phase (3) images or the real (1) and imaginary (2) images (which you would use as the input for Fugue).I note this is a "Work In Progress" so perhaps you can look at the image on the console to determine what type of image this is. In general I prefer my software to exit with an error than generate a bogus dataset. I think you have a few options:
 1. Can you export the data as DICOM? That should give a nice result, and will provide better archival data than a proprietary fromat.
 2. You could patch dcm2niix to handle these (and issue a pull request to share your solution). 
 3. You could edit the PAR file with a text editor to fix the problem directly
 4. You could send me a sample and I will see if there is a solution.