I've been wondering... is a strategy built around making fairly strong increase then winnowing (by requeening/uniting/winter losses) … anything less than successful treatment free beekeeping?
I'd argue that as long as care is taken to allow true assays of resistance to be made, by not providing artificial support by creating brood breaks while making increase, and by allowing colonies to live to maturity, taking ideally 3 years thriving as evidence of strong levels of resistance... this amounts to treatment free beekeeping of a healthy and sustainable sort - even though replacement rates might be of the order of, say 50%.
Just because a fairly high percentage are terminated - one way or another - that doesn't make the apiary a failure in either beekeeping or non-treatment terms. In nature many colonies fail; and in traditional beekeeping strong re-queening and losses of the order of 10 or 15% were considered normal.
To be continued... message limit for some reason...
Perhaps one of the the best questions to ask is: does such an apiary weaken local wild/feral bees? I don't think so - in fact the opposite is probably true.
Second question: Is trying to avoid that system - and those sorts of losses - automatically asking for trouble regardless of varroa? It seems to me that aiming to maximise productivity by systematic requeening and treatments ('monkey'-beekeeping) systemises the use of treatments - unless steps are taken to avoid that, and generally creates more problems.
What I'm trying to get at is: the best that beekeeping can be in a rounded evaluation might be one in which losses are expected and covered. That might be less than optimimum in productivity terms - though probably not very much less.
And it means, if we accept it, that evaluating success by proportionate losses is not useful. I think that is what I'm trying to put my finger on here.