Not a doctor, but from what I've read there are quite a lot of problems in the way:
- Lead time bias. If you discover cancers early and patients survive more, is it because you're treating better or because you discovered (and started counting) earlier?
- False positives and negatives. All tests will have them, and for low prevalence cancers it might mean that testing doesn't give too much information and just increases load on healthcare systems and puts patients through unnecessary tests and procedures.
- Indolent diseases. Extensive testing might discover tumors that don't have symptoms and are not life-threatening, so you might treat people (with the risks associated) that didn't need to be treated.
And that's if the tests themselves can be developed. Apparently there are a lot of types of cancer, each with different markers, behaviors, affecting different cells. It doesn't seem likely that a single test will be able to catch cancer in general.