As I recall, they did not take the error rate into account. The result was dashed by somebody looking at their error rate and saying "we can compute that distr

As I recall, they did not take the error rate into account. The result was dashed by somebody looking at their error rate and saying "we can compute that distribution to within 5%* with a classical computer." One would hope that they learned from this, but historically, supremacy claims have been rather short-lived. * number pulled out of a hat