Culling marginal workers during recession leads to greater per capita productivity?

In a recent Econ Talk podcast, host Russ Roberts and guest Garett Jones discussed the relationship between an economy's health and worker productivity, pointing to the real business cycle theory introduced in the early eighties by Edward Prescott and Finn Kydland. RBC theory predicts that during economic slowdowns, productivity will drop because innovation is sparse (it is this real lack of innovation and of attractive investment opportunities that is said to be the cause of the recession, rather than monetary policy). In the last three major recessions, though, productivity has increased during recessions and dropped during booms. The latter pattern seems commonsensical to me, and I was unaware that in previous recessions the reverse had been observed. Jones goes about describing the putatively conventional explanation for why productivity (at least over the last couple of decades) goes up during downturns:

During a recession is when the boss comes and tells all the workers, "Hey, I'm going to make you work harder." ... There's certainly some of that going on, on some level--bosses do love to squeeze the last bit of blood out of a stone. ... There's this point that even Marx made about the army of the unemployed: The boss can point to all the people out on the street and say, "Hey, those people want your job, so you better work harder for less money."

He and Roberts also point to substantial reductions in R&D and marketing spending by companies during slow periods, while budgets for physical production tend to be reduced more modestly, as another reason for increased productivity. Since productivity is measured by dividing worker hours by actual production, the relative reduction of indirect support functions bolsters productivity numbers.

What the two George Mason economists didn't discuss, and I also did not hear mentioned last week when labor productivity figures were released, were any effects resulting from the culling of marginal employees. If the number of consulting slots stagnate and then begin to decline, either per capita workloads are reduced, or the number of employees is reduced. To the extent that anti-merit obstacles like tenure or demographic considerations can be circumvented, the latter provides an opportunity to increase productivity (either out of genuine necessity, or by providing cover for what needed to occur anyway).

I am not aware of any hard numbers on the various elements of productivity changes during the current recession--it may loom large in my eyes because I've experienced it on multiple occasions in the last two years (on the delivering end--my plate is still full, although it is a wrenching process to go through even when you are giving the boot), and because I'm not in the manufacturing sector, where an individual employee's performance is less determinative of his fate. Or it could be the strong strain of HBDism in my blood that is skeptical of the emphasis placed on the value of motivating and driving people to be better performers at the expense of focusing on finding the most capable people in the first place (that is, who you are is the most relevant question an employer should be seeking the answer to, not what ways to squeeze the most blood out of the stones they've already collected).