I don't know much about TFT's nor the SSD1963 so this is really a shot in the dark, but if it's anything like the SSD1289 then you might be initializing it incorrectly. Take a look at page 71 in the SSD1289 datasheet (http://www.kosmodrom.com.ua/el/STM32-TFT/SSD1289.pdf). It shows that you can use 262k and 65k color in 16-bit mode. Perhaps you're initializing it in one of the 262k color modes? That way the display would receive what it thinks is the red and green data, then presumably you would send a command to move to the next pixel and the blue data would never be received. Seeing as the library already had other errors (the rotation) this could be a likely possibility. I'm not sure about the whole B,G,R vs. R,G,B though.