Hi Soichi,
While there are many mux overlaps between the Flexbus and NAND, it would be possible to only use the lower 16-bits of the Flexbus which are not muxed with the NAND pins. And as long as you aren't using some of the optional Flexbus pin options (like TSIZn, BEn, etc), sharing the NAND and Flexbus might be possible in your application. The TWR-LCD only uses FB_RW and FB_CS for instance, so depending on what you need for your FPGA, this work-around may work for you.
The main thing to keep in consideration is that FB_OE and NF_RE are muxed on the same pin, and that is a commonly used signal on the MRAM and other devices.
The reason the Flexbus wasn't brought out on the Vybrid tower board was because those lower Flexbus pins are muxed with the QuadSPI pins, and that was a higher priority on the tower board to preserve signal quality.
-Anthony