You also seem to believe there is a black hole at DP where 1 out of 3 books fall into, never to emerge. This is a patent fallacy.
The fallacy is in assuming that the only way DP can waste volunteer efforts is to never ship some particular book. On the contrary large and increasing queue sizes can as effectively waste volunteer efforts as much as never shipping some particular book. Again, consider the Russian Roulette test: DP managers randomly shoot 1/3 of the projects at DP (prior to PP). How do these murders affect the shipping rate out of DP? Answer: They don't change the shipping rate out of DP. Conclusion: If you can destroy 1/3 of the projects at DP without affecting the productivity rate out of DP then 1/3 of the productivity at DP is being wasted. How is that productivity being wasted? By sticking it on large and increasing queues. Consider a factory that only ships 2/3rds of what it ever starts to make. Does the unfinished inventory represent value or not? Well, the factory only *realizes* value by shipping product. The shipped product has value, and eventually every piece of product gets shipped, but as long as the factory only ships 2/3rds of everything it ever makes the fact remains that the cost of manufacturing is 50% higher than it need be. IE the factory is only running at 2/3rds of its potential productivity. That unfinished inventory *might* be considered to have value, but only if new owners buy out the old owners, and change the manufacturing process such that you don't have unshipped inventory plugging up the factory anymore. Or if buyers get tired of paying 50% more for products than they should be and stop buying, then the factory has an opportunity to work off that unfinished inventory, realizing its value -- assuming they can lure back buyers at the new now lower price that doesn't include the wasted 50% markup for product started but not yet shipped. In the DP case what this analogy means is that DP gets a chance to work off the inventory if and when P1s get tired of DP wasting their time and energy and thus stop putting new work into the head of the DP queue. But DP needs P1s since they represent the future of DP. Now how can it be that a factory only ships 2/3rds of what it makes but at the same time it eventually ships every item? Consider for simplicity that the factory makes rolls of toilet paper and ships those rolls out to customers based on a "First In First Out" FIFO toilet paper roll queuing system. Does every roll of toilet paper eventually get shipped? Yes. But the problem is is that the queues are constantly getting larger, and as they do so they consume 1/3rd of the factory's resources. Consider if we changed to a "Last In First Out" queuing system. Does that change the nature of the problem? NO -- a roll of toilet paper is a roll of toilet paper. But now, based on LIFO it becomes obvious that some rolls of paper never do get shipped -- the 1/3rd of the older toilet paper rolls at any given time never get shipped -- 1/3 of all toilet paper rolls every made, and the situation keeps getting worse. But the choice of FIFO vs. LIFO queuing system in no way changes the nature of the problem -- a toilet paper roll is a toilet paper roll. Thus, on the contrary to the previously stated hypothesis, it is NOT necessary to have a "black hole" in order to waste time and effort. All that is necessary is to have a large and increasing queuing system -- whether that queuing system is LIFO or FIFO. Or stated another way, large queuing systems ARE the black hole. The mere fact that any given book eventually makes it out of the queue is not sufficient to keep the large queuing systems from being a black hole -- as long as the black hole continues to suck in more than it spits out.