So this is true, and these are both points that are often missed when discussing these issues. But at the same time, it is worth noting that A) a variety of the studies listed were with people rather than animal models and B) while cancer is better thought of as a collection of distinct diseases, that doesn't stop there from being treatments that work for a large variety of different cancers. For instance, many forms of chemotherapy target cells which are dividing, which since that's a common property of cancers, often manages to hit on a large number of cancer types. Similarly, we can classify cancers in some detail, and large classes of cancers often have the same or similar properties.