I think this thread is missing some significant points.
Evolution, through selective pressure, is less about the "form" of the pressure than it is about the "strength" of the pressure. The greater the pressure (in any form), the faster and more strongly selection occurs. The resulting changes are "evolution."
The selection can either be natural or artificial.
Secondly, the source of the selective pressure is important. Not the form; the source.
Let's just take a blood-sucking insect, a mosquito, on a human as an example.
The mosquito bites the human and withdraws enough blood to fill itself. Selective pressure on the human? Very, very low to none. The amount of blood withdrawn is highly unlikely to pose any threat to the human. The mosquito, and, in most cases, the allergic reaction to the mosquito bite, however, make humans desire to avoid the mosquito. So, humans have developed a defense. Slap.
Now, if our human misses a mosquito, well, no real harm done. The mosqito escapes with her life and a full meal of blood. The human survives quite nicely, and the mosquito bite confers no real loss of fitness on the human.
So, the selective pressure from the mosqito on the human is negligible, at best.
But, if that mosquito is carrying a disease -- let's just say, malaria -- the human's life is at risk because of the malaria. Loss of life, if occurring before offspring can be left, seriously reduces evolutionary fitness. So, the selective pressure is much, much higher on the human. But not from the mosquito; from the plasmodium that causes malaria. As the chances of contracting malaria increase, the selective pressure increases.
Now, a "reasoned" method of reducing the selective pressure might be to reduce the population of mosquitoes that carry malaria, or a way to reduce the incidence of mosquito bites, but selective pressures don't function that way. The selective pressure is from the plasmodium, not from the mosquito.
Same thing goes, seems to me, for Varroa on honey bees. As long as the viruses or other pathogens carried by Varroa are the source of the selective pressure on honey bees, the adaptations that will enhance the evolutionary fitness (survival) of the bees will be in response to the direct source of the selective pressure.
Same ideas apply to resistance to "soft" treatments developing among Varroa. As long as the selective pressure is low (either the mites face relatively little chance of exposure to selective pressure, or the pressure does not significantly reduce the evolutionary fitness of the mites), resistance will develop very, very slowly -- if ever. But if the selective pressure increases, the speed with which resistance (or adaptation) will develop will increase correspondingly.
Just an interesting note on mites overcoming resistance: T. L. Harvey, T. J. Martin and D. L. Seifers at Kansas State University used experimental pressure to measure the speed with which wheat curl mites (Aceria tosichilla) can overcome plant resistance to the mites. The plant resistance is frequently physical, not chemical, such as "hairy" plants preventing the mites from acquiring feeding sites. Now, these mites are not the same as Varroa, but they are both mites. The wheat curl mites were not given a chance to switch plants -- either they managed to feed on the plants presented to them, or they starved to death. The mites overcame the best forms of host plant resistance available at the time of the study in 60 days. More interestingly, 60 days represents 8 generations of wheat curl mites.