This is all a conundrum for WCG for the last two years now, and yes, to some degree it is a problem that most of the BOINC projects are dealing with.
As far as WCG is concerned, from a "regular" participant's POV, it is now, 22 month after the restart under Krembil/UHN, clear that they bit of more than they could chew.
They are short on funds, which wasn't helped that some part of the infrastructure/storage took a deep dive and needed to be replaced. They also have refused any help with the system admin stuff, despite being short staffed, with the rational that in their (healthcare based) setup, they can't just grant access to "everyone".

That has lead to the fact in those 22 months, we barely had 4 weeks of really uninterrupted operations, despite the powers at WCG Towers proclaiming repeatedly "full restart" and now even starting competitions, while the system might choke at any point again. Just the last few days, about 4 months after the problem started, the purging of already validated WUs seems to be progress at a pace that one can finally get the Results page loaded without repeated retries. And we are still waiting for those "lost results" from the initial start of WCG (May 2022 through late September 2022).

All the while, there are some "power users" that are demanding that THEIR requirements are being met, million years badges, endless supplies of WUs FOR THEM, etc.
Yes, as far as the topic of all the projects at WCG, there are for a lot of people more worthy causes than finding new prime numbers or possibly contacts by ET (or some bogus crypto stuff). And the original goal of BOINC was to use "spare" computing power, while a good number of participants seem more content about setting up crunching farms and become point w****s, all the while new participants, who want to help with cancer/COVID/etc research are getting disappointed because there are no WUs available or the system took a dive again.
Yes, looking at it from plain numbers game POV, one participant crunching 5000 WUs in the time 100 "normal" users each do 50, for the goals of the project, that seems to be the same. But how many of those "power users" are going to leave once things are not going smooth, taking all their processing power with them? Or move on to greener pastures that "pay better". While at the same time, a lot of Joe Average users have left or never got a chance to participate in the long run...

As I see it, WCG needs to get first and foremost into a state where there is a smooth, uninterrupted operation possible. That simply is not the case at the current time. And IMHO, it doesn't make any sense to either increase any WU limits or start any competitions, which only serve a short term goal (and might quickly overwhelm the system, as we have repeatedly seen in the past 22 months) before a REAL stable state has been achieved. Instead of just wishful thinking...


Ralf

Al, The "you" was generic. Sorry for the confusion. I agree with you (that is you specifically) that any contributor that downloads so much work that it isn't returned by the deadline is a detriment to a project. Where the whole discussion got started yesterday with David Anderson was in connection with the 1000 hard-coded limit in the client. Participants with high-core count processors requested the ability to download enough work to at least give a one day cache (we know what it like here at WCG) but we were dismissed out of hand. There was an incident opened in github that had been open since 2019 and nothing done about it. It's a simple one line change so I asked why it had been open for 4 years. I can only assume it was a pocket veto. Anyway, David sent what I considered to be a canned response, because it was essentially the same response I got back in 2015.