In blood cell cancers, Steve McCarroll's group has shown that oncogenic somatic DNA changes are readily discovered through routine sequencing, and these are prognostic of poor outcome.[1] If you do this on a pre-disease schedule based on the risk estimates from your first round of sequencing, you will be able to detect blood cell cancers at any stage you like.
In non-blood cell cancers (say, breast cancer), you don't expect to capture any of the tissue-specific cell of interest from a blood draw to determine if there are somatic mutations occurring. However, there is growing evidence that there is often enough circulating tumor DNA in the blood that it can be of diagnostic significance.[2]
In contrast, for diseases that do not involve somatic mutations (in other words, for diseases aside from cancer), an RNA-based assay could be quite interesting. For example, getting better characterization around non-dilated cardiomyopathies would be of broad interest.