Works great on STM32L4

[Defragster]

I just hooked this up to a Tindie_Tlera Ladybug and it is 18 times faster over a test suite that goes from 18323.891 ms down to 946.418 ms for a combined set of 18 tests - most just POP the image onto the screen! This is on an 80 MHz STM. The tests look like they run the same - but too fast to tell :)

It has room for optimization with #ifdef. Wondering (hoping) if you are back online Sumotoy and would take a PULL request if I got the changes done right?

Small things like it has twin SPI, can do 16 bit writes for color - and should have direct port writes to replace digitalWrite()'s that I didn't test yet. There may be others.

[laurentopia]

This would be very useful for esp32 as well