If you only look at everything that has already passed the test then sure the test does not seem important.
The set of all possible cancer treatments however are not just the set of treatments that pass animal models. Drugs etc need to go though several hoops before they are even tested in mice, and even more fail in mice. Everything that’s made it to this stage is extremely promising relative to the baseline.
About 1/3 of drugs that show promise on mouse models are ever tested in humans. After that Oncology has a 8.3 percent success rate for 2015. So, when looking at ‘cured cancer in mice’ something like 2.5% is about the right ballpark, which I would consider promising as a general baseline.
For some comparisons the success rate in clinical trials averages around 14% though it ranges from around 33.4 percent in vaccines for infectious diseases to 3.4 depending on the type of treatment being tested. https://www.centerwatch.com/cwweekly/2018/02/05/new-mit-stud...
Baseline is harder to calculate, but 10^-8 is probably reasonable depending on what’s considered.
> About 1/3 of drugs that show promise on mouse models are ever tested in humans.
Do you have a citation? This seems totally off to me. In so far as the number refers to something real, I suspect "shows promise" must mean something much stricter than the threshold that this particular treatment has passed. That is, I would bet against a 2.5% chance that this particular study leads to an actual drug approved to treat pancreatic cancer.
The set of all possible cancer treatments however are not just the set of treatments that pass animal models. Drugs etc need to go though several hoops before they are even tested in mice, and even more fail in mice. Everything that’s made it to this stage is extremely promising relative to the baseline.