Several factors come to mind. As Wayne points out the moving mass is an issue, sagging in the gap, resistance of the traces (planar implementation), width of the ribbon (wants to be wider for power handling at low end of band but narrow for wide dispersion and frequency coverage across upper end of band), intensity of magnetic field (neo magnets likely will help this sitaution), etc. What often results (B-G 75" planars come to mind) are implementations that have low sensitivity (less than 90 dB SPL), compromised upper end (above 15 kHz), limited dynamic capability/power handling on the low end of the band (below 300-400 Hz), etc. I do like discrete tweeter line arrays made from the 5-7" true ribbons from Fountek and Aurum Cantus. You get excellent horizontal dipersion, high sensitivity, and exceptional sound albeit at a price and coverage only down to about the 1000 kHz area .
Jim