tests: increase timeout duration when awaiting a timeline update

[bnjbvr]

Fixes #3931, or so I think: the test waits for a stable state where there aren't any new updates to the timeline, before trying to edit a timeline item, and observe a precise sequence of timeline updates after the item has been edited.

The failures showed that the stable state wasn't reached, because in some occasions, we'd receive the original text as a new timeline update — which could also be the sign of a failing server / duplicated event or other things, but I believe it is just that 1 second timeouts ain't enough for code running under CodeCov.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 84.24%. Comparing base (abbe2ec) to head (19400fc).
Report is 4 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4005      +/-   ##
==========================================
- Coverage   84.25%   84.24%   -0.01%     
==========================================
  Files         266      266              
  Lines       28328    28328              
==========================================
- Hits        23867    23865       -2     
- Misses       4461     4463       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[andybalaam]

I agree: we could easily see some kind of system slowdown that takes longer than 1 second, so this change is good even if it doesn't fix the problem.