After some testing, the nRF52832 on the Feather module being used is not going to work well for this project alone.

The data is fed into the LCD approximately every 14us:

However the critical radio sections of the nRF52832 can take much much longer than this:

These Bluetooth critical sections have the highest priority on the nRF52832 (otherwise they wouldn't work), so several LCD data events can occur before the LCD data ready interrupt is actually triggered (aliasing/skipping data).

So now to find another method of gathering the LCD data...