Did you mention that according to Nixtla study Chronos underperformed simple statistical ensemble?

According to the results of the study “Amazon's new foundational model Chronos is 10% less accurate and 500% slower than training classical statistical models.” when tested against a number of datasets.

TIDE is a good model, but one can’t make conclusions about Chronos by testing it against one model. I agree that Chronos is much better than Lag Llama though.