"Balance of costs and expected benefits" is definitely more accurate. For example, if the claims for this drug are accurate, I'd be surprised if it didn't make the cut because it has some pretty clear long-term benefits. The treatments you hear about the NHS not funding are generally ones which are both expensive and not terribly effective.
Indeed: The typical examples are cancer drugs for a specific cancer, which give a few extra months of good-quality life. The cost/benefit on these is not clear-cut.