[maxander]

Therapies that could reliably cure any cancer are often worthy of research, but I'd put the expected value of these approaches lower than that of "ugly morass" standard research.

Reason 1; cancer therapies have to effectively kill cancer cells and not important non-cancer cells. Cancer therapies that exist today achieve this by mucking about in mechanisms that are important to cancers and not (or not so much) to other cells. But cancers are diverse- there may just not be anything that's common to all cancer and not shared by healthy cells. Telomere regeneration is a nice candidate, but as you noted, its also important to stem cells- can humans withstand enough damage to their stem cell populations to make a meaningful difference to difficult cancers? (For that matter, telomere regeneration is also important to gametes, so telomere-based therapies may also render the patient infertile.)

Which means we're likely to be stuck dealing with a dozen or hundred-odd different treatments, each of which targets some eccentricity of a specific cancer subtype, but-

Reason 2: There are lots of smart people plugging away at this problem, backed by lots of research money. In addition to being an objectively important problem for humanity, cancer is also an interesting biological conundrum and also the most likely cause of death for a bunch of old people with stratospheric net worths. Even hacking at the problem in bits and pieces, its not unlikely that cancer will become a completely tractable disease (even if a golden-bullet "cure" doesn't arrive) within our lifetimes.