Before they lose all credibility?

They lost that when they shut down the old server before the new one was fully
tested

I agree. Setting up a script to hammer the server every 10 seconds doesn't help the problem, it creates a new one. Hopefully people are wise enough not to go down that road.

The problem is that the only 2 machines I have not running the script at this time are running dry because of incomplete download during days.
The whole situation is totally creepy.
What I don't understand is the reason why Krembil is unable to solve the problem in a timely manner.
The problem is so huge and constantly occurring that it should not be as difficult to identify it and to search for the root cause(s) and to solve it.
Is the Krembil staff finally able to perform failure analysis and to conduct investigation?
I am fully aware that IT can be very complex and failure solving can be demanding. However regarding WCG we speak about BIG failures permanently occurring since weekS / months.
In an industrial context it would be simply impossible!
The "hard WCG core" is willing to help, to actively contribute to failure investigation, to share knowledge ... instead of leveraging this willingness, members are ignored and they have to pay electricity bill for nothing.

In June 2020, one of my hosts experienced a very strange issue. After reporting the trouble, I had direct contact with Keith for performing several tests until we solved the problem. It occurred very rarely and it was accordingly relatively difficult to reproduce. Nevertheless after a couple of tests, validating step-by-step the root cause identification and afterwards the solution, we were able to solve the problem within less than 10 days, mostly because I was not full-time available for performing the tests faster.

We expect such an openess and a willingness to solve TOGETHER the problems.
Cheers,
Yves

This evening unlike the last three weeks, downloads from Krembil's version of WCG are working a little bit better. A little bit. Most downloads of the MCM fail at least three or six times after showing 107 bytes downloaded and then -- download failed. Connected, downloaded 107 bytes ou of 3--hundred bytes, failed, retry in an hour? WTF
Bigger downloads, when they don't fail on the first few bytes, which mostly they fail, then the downloads succeed
IBM left Krembil with a serious problem. I'm a senile (elderly) person who sometimes hits the retry again :) (and again and again each minute) button on the downloads. Some people might think that is looking like paint dry.
And the downloads are O so slow, better than dial-up, but so slow, I have a lously 50Mibi adsl but downloads only about 50 Kibi:?
not good
Did IBM dump WCG on Krembil with no hardware or software support? Kinda looks like it - but I dunno,.
whatever, I keep downloading (super slow) and supporting WCG

Yup, it is a great annoyance.
"What I don't understand is the reason why Krembil is unable to solve the problem in a timely manner."

Did IBM dump WCG on Krembil? with no hardware of software support?? Just asking :)
Duh :)

I do own a very few shares of IBM, not enough to be heard at the next AGM.