More accurate than lightning: Oxford's new quantum gate

the article says you're more likely to be struck by lightning than for this gate to fail. The accuracy is a record-breaking 1-in-6.7 million operations.

But this is for a single-qubit gate. The article claims that the best two-qubit gates still have error rates of around 1-in-2000.

My question is: how much does this incredible single-qubit improvement change thinks, given that two-qubit gate errors are still a major problem? Does a breakthrough like this directly translate to improving two-qubit operations, or are they completely separate challenges?