And the Parkway is a continuous descent and the Canarsie is a step-down to 800'
So in good weather, fly the VOR approach with the profile of the visual.
Approach plate of the parkway in a better scale with more visual cues as well..
EK created FMS coding from the MAP to the threshold after the Canarsie approach. Very useful and you can make it a continuous descent as well, but it may create the false impression that you don't need to be visual until short final. The minima is still 800' and you don't continue unless visual with the runway or lead-in lights. The whole FMS coding to the runway is still a visual segment, which might have been flown on instruments in this case.
Another possibility is that with all the different approaches created by EK, they might have selected the wrong one and/or the wrong minima. The RNAV Visual requires a minima of 300' to be entered in the FMS according the EK procedures. Wrong in my opinion, because this can also create the false impression that you don't need to be visual until 300' on final. It is still a visual approach, but full automation is standard at EK.
If they selected the RNAV Visual instead of the Canarsie approach, selected 300' as minima and thinking that they were flying an instrument approach, then that might explain why they ended up so low.