The most fundamental system we have to quantify the importance of scientific research is broken at its core, a new study reveals – and all it took was a single punctuation mark.