We could write to buffers at 60 Hz effortlessly with computers from 1999, speeds have increased more than enough to write to buffers at 120 Hz and more, even with 16x more pixels.
1/120th of a second is a huge amount of time in CPU/GPU clock ticks, more than enough to compute a frame and write it to a double buffer to swap, and more threads should make that easier to do, not harder: more threads can compute pixels so pixels can be put in the buffer faster.
If there's problems with connector standards, software side of things, multithreading making it require third-party complexity, then that's a problem of those connector standards, the software, things like the LCD monitors themselves trying to be too smart and add delay, etc... Take also for example the AI upscaling done in NVidia cards now: adding yet more latency (since it needs multiple frames to compute this) and complexity (and I've seen it create artefacts too, then I'd rather just have a predictable bicubic or lanczos upscaling).
Same with audio: why do people tolerate such latency with bluetooth audio? Aptx had much less latency but the latest headphones don't support it anymore, only huge delay.