"I can't draw a conclusion from this data point, therefore it's useless data" is something no researcher has ever said. If you can't draw a conclusion you get more data. And here's the thing, right—getting more data isn't Mathilde's job. It's not her responsibility. It's her job to pass on that data to someone else who does have that responsibility.
"A powerful and trusted wizard had a miscast, weakening reality in that location, and through that weak spot an entity self-identifying as a Lord of Change told them that an unknown number of Chaos Champions have begun competing for the title of the Thirteenth Everchosen" is a valuable data point to someone within the Empire's intelligence apparatus.
For a start, it tells them that a champion hasn't been chosen yet, and it allows them assign context to the actions of known chaos threats. Two cults tearing each other apart is no longer the "inherent self-destructive nature of Chaos" but two candidates attempting to subdue each other for the favour of their Gods. A raid between Kurgan tribes witnessed by a passing merchant might be about cattle or silver or slaves—or it might be a sign that a powerful leader is attempting unite warriors under their banner through a series of quick military victories, and when they reach a tipping point they are going to start raiding into Kislev.
And if the frequency of reports about the activities of champions, and portents about their behaviour, rises, then that suggests they are becoming bolder, and maybe more desperate, as they receive more blessings and the number of competitors goes down.