Why it spread everywhere

LoRA democratized fine-tuning. Suddenly a hobbyist with a single GPU, not just a big lab, could specialize a large model. It is now one of the most widely used techniques in the field, and it underpins the huge ecosystem of community-made model variants. Whenever you hear about someone "fine-tuning" an open model at home, they are almost always using LoRA or a close relative of it.