Having tasks "ready to report" isn't a problem. To minimize the traffic to the scheduling-server, and more specifically the database-load, the BOINC-client intentionally waits until there's another reason to connect than just having a task "ready to report".

The most common reason for scheduler-connection is to ask for more work. If there's no reason to ask for more work, either due to user decreasing cache-size, or due to estimates was too low so DCF has increased a lot, it can take some time before needs to re-fill with more work. Well, of course, if you've set "no new work" you won't ask for any either.

But, rest assured, client will at the latest report 24 hours after a task was marked "ready to report". There's other reasons like if less than 24 hours until deadline and so on, many of the rules is mentioned in one of the FAQ's.


BTW, using the method rilian mentioned of setting "connected..." to zero won't report tasks any faster, it can actually have the opposite effect, since one of the rules for reporting is:
If less than "connected..." days until deadline, report finished work.

Setting a small "connect..." did report faster before, but this meaning was deprecated in BOINC v5.10.14.

This is not abnormal at all if the cache of your quad is set to about one day or more.On mine I have seen up to 4 "ready to report" tasks with a cache of only 0.2 day.
Anyway for a 4-thread machine 20 HCC WUs is the work of less than 6 hours, or even much less if yours is a fast one.

Edit: Marti, if I had seen your 6-day cache info before pressing the "post" button I would have replaced "is not abnormal" by "is peanuts".
The "bandoneon" effect described by Ingleside is amplified by the number of WUs waiting in the queue. When the estimated duration of your HCC WUs in queue (probably 600 or more) is readjusted by 10 % more it is the same as if you had had 10 % more WUs of the previous estimated duration in the queue, i.e. 60 more in my assumption.