One can simply have a scenario in which farming disrupts the viability of hunter-gathering by disrupting the local ecology enough to make most people who adhere to that approach die of starvation or dramatically decline in numbers due to factors like infant mortality increases or farm culture disease, and to have a minority of women taken as brides (at least some of the time in arrangements as voluntarily as would be the case with couples within hunter-gatherer society) who leave legacies. The new wave doesn’t have to be the direct warlike means of slaughter of the men. Indeed, it is hard to believe that warlike slaughter could have been that completely successful at that relatively rudimentary level of military prowess.

There might have been selective slaughter’s of men, but exile to marginal hunting and foraging areas seems more plausibe as the main basis for the marginalization of the Y-DNA of members of the hunter-foraging societies.

Absence of R1b may be due to their presence elsewhere, such as in major river valley settlements colonised earlier, and in possible hybridised forager and semi-forager populations. So “late foragers” may actually turn out to have plenty R1b (if late foragers are found with some R1b as I think it’s likely, it’ll be interesting to see what people will say). New arrivals such as some secondary waves may have had enough agricultural advantages to have a bigger demographic impact in less fertile regions left fallow by earlier waves.
R1b could then regain ground if it was based in the richer areas, which upon acquiring innovations, would due to higher density and surplus be natural elite-formation centres for less fertile regions over several millennia. These events wouldn’t necessarily imply major autossomal changes or discontinuity. Patriarchal societies would tend to transmit Y-DNA (prestige of uniterrupted male lines to elites) but not so much autossomes.
Greater amounts of forager-associated mtDNA than found today also can be explained by Treilles being a new settlement by a new Near Eastern wave, in a previously fallow area.

Unfortunately for Sykes, he was just plain wrong. Now that aDNA researchers have better techniques to guard against contamination, and can test to a much higher degree of resolution on the mtDNA tree, the picture has turned completely around. Two papers last year showed that Neolithic people in Northern Europe were completely unrelated to their Mesolithic predecessors, who were overwhelmingly U4 and U5. It no longer looks as though H1 and H3 were spread from the Franco-Cantabrian refuge either. See my blog post: Franco-Cantabrian refuge for what? Not mtDNA H1, H3.

New papers attempting to set their own material in context do not always reflect the latest discoveries. There is often a long delay between submission and publication of a paper. This is the second paper which has attempted to reconcile the old Sykes story with what is emerging from Y-DNA, not realizing that the Seven Daughters saga has gone the way of the dodo.

Some of the mtDNA haplogroups found in this study are indeed ancient in Europe – the U5 for sure. But not the majority.

Since similar behaviors have been seen in various primates (and, IIRC some other mammals and perhaps non-mammals) it seems to be a social extrapolation of a behavior related to evolutionary fitness.