You need to take a step back from these testing modalities and think about how they are calculated. Screening for disease has two competing arms, which is slightly cynical of me to say but consider it like this. The accuracy and benefit of the test, is competing against the interests of the government in wanting to subsidise them.
What this eventually comes down to, is cut off limits. For example, screening for bowel cancer is nationally subsidised, is cheap, very sensitive, yet not very specific, as lots of things can cause blood in stool, such as a minor haemorrhoid. However, despite this patients with a positive result are referred for a colonoscopy because the disease is so so so treatable if caught early. So there is a huge survivor benefit
But we stop screening for it at 74 years of age because the benefit of anything we found at that age, would be less than the survival benefit of treating whatever is found when compared to standard life expectancy.
On the other hand you could have VERY specific tests, that are not very sensitive, such as various blood tests for autoimmune diseases. You conduct the blood test and if they come back positive, that person (near) absolutely has the disease, yet a negative result doesn't exclude disease.
A good screening test has some characteristics, it needs to be tolerable. If a patient wont undergo the testing, then it is useless. Examples of this might include transvaginal ultrasound for ovarian cancers, which very large, randomised and double blinded studies have demonstrated that women will undergo routinely, because ovarian cancer is nasty, hard to find and harder to treat. . Yet despite these large studies, we do not have routine screening for ovarian cancer as the studies demonstrated that despite doing serial ultrasounds, blood tests for ovarian tumour markers and follow ups, there was no survivor benefits to the whole process. So this remains an unmet need within health.
A good screening test needs to be applicable to a population. So a blood test fits this descriptor.
A good screening test needs to be accessible, this is where cost comes in, as does the level of subsidy a government is willing to give.
The level of subsidy that a government will give really comes down to, what do they get out of it. They will ask, does the cost of administering this test to a population, equate to equal or less than the cost of treating it now or later. . So they would need to compare any new test, against current test. AKA mammography.
Specifically in regards to the numbers given. Sensitivity and specificity are inverse relationships. So its really hard to define a perfect test. But at 85% specific, some 15% of patients told they have a negative test, actually could have it.... not good. So from a screening point of view, I would personally like to visualise the tissue. After all, a lump is a lump.
Lastly, if folks care to look into this. This below table is how these things are calculated. A quick google of 'the perfect screening test' or similar will give more than a days worth of reading.