Looking at DVB-T2 technical specifications I see that sub-carriers spacing has, in the worst case, a value of few hundred hertz, (something like 279 Hz). 32K sub-carriers are used to reach such a low value. I suppose that complex PAPR reduction algorithms plus strong constraints for carrier frequency stability are needed to deal with this number of sub-carriers and frequency spacing. On the other hand, I don't expect that the channel's frequency response would change significantly over few hundreds hertz.
What are the reasons why such a high value of sub-carriers is used?