If this argument were sound, tailless designs (i.e. without a separate horizontal stabilizer either before or behind) would be optimal, but what matters to efficiency is the overall lift-to-drag ratio within all the other feasibility constraints. No-one obsesses over lift-to-drag ratios more than sailplane designers and their customers, and the fact that the medium- to high-performing sailplanes are neither tailless nor canards, despite there being airworthy examples of both canard and tailless gliders, is telling us something, at least in the range of Reynolds numbers relevant to glider flight.
Three issues with pusher-prop "tail-less" designs:
1) While they are more stable in nominal flight regimes, they are far harder to recover to stable flight from perturbations. It turns out, it's (overall) much safer to have a plane that will BOTH stall easier (more predictably) and recover easier than one that is less likely to stall in the first place, but difficult to recover from. One analogy I often use is the difference between a mid-engine car and a front-engine layout. While the mid-engine car has a greater overall theoretical "handling" performance ceiling, a front-engine car behaves more predictably (less twitchy) at the limits.
2) They are more susceptible to CG/balance issues so they have less practical cargo capacity because just a weee bit of pitch/yaw/roll trim results in a drastic drop-off in the aforementioned stellar lift efficiency.
3) They have much longer take-off and landing runway requirements due to less ground-effect and much less overall wing efficiency at near-stall speeds.
Are you saying that having a horizontal stabilizer is not a source of inefficiency? This isn't an argument, it's a basic fact about airplane design. They necessarily contribute to overall drag.
In practical airplane design, there are considerations other than drag that make horizontal stabilizers a worthwhile compromise.
Firstly, 'having a horizontal stabilizer is a source of inefficiency' is not an argument, it is a fact (one that might be a premise in an argument, but see below.) As you know this, how did you get from seeing that I said a certain argument is unsound to supposing that I am disputing this fact?
Maybe you think it is the only way that argument could be unsound, which brings us to the second point: the argument I am commenting on is not 'having a horizontal stabilizer is a source of inefficiency, therefore canards are more efficient', which would not even be valid. It is, instead, the argument that canards are more efficient because the conventional horizontal stabilizer usually produces a downwards force (incidentally, this is not always so [1].) While this may seem an obvious conclusion at first sight, it tacitly presumes a sharp separation of concerns which does not hold in practice.
The argument 'having a horizontal stabilizer is a source of inefficiency, therefore tailless designs have greater efficiency' (which was not made in the post I was replying to, but which is sometimes alluded to) does not hold up any better, on account of the compromises in making a stable and controllable tailless airplane (at least without active stability augmentation.)
[1] For some conventional airplanes, with the GofG near its aft limit, the horizontal stabilizer will produce an upwards force at low speeds (without being unstable as a consequence.) This happens to be the case for many gliders. A while back (and possibly now lost - at least, I have not been able to find it), there was an interesting article (by Wilhelm Dirks - co-founder of DG Aviation, I believe) explaining why, in practice, this cannot be exploited to get a little bit more performance out of a sailplane.
