Fix: Multi-window UI tests can produce inconsistent results

After creating a new scene in a separate window when performing UI
tests, the respective view layer for the window may not be updated
immediately in the event loop.

Previously, this was mitigated with a single `yield` statement that
would delay processing by a single tick. To fix this issue, this commit
adds the capability to yield for a specific `timedelta` and waits this
amount of time for the two affected tests.

Pull Request: https://projects.blender.org/blender/blender/pulls/136012

tests/python/ui_simulate/modules/easy_keys.py

@@ -2,6 +2,7 @@
 #
 # SPDX-License-Identifier: GPL-2.0-or-later
 
+import datetime
 import string
 import bpy
 event_types = tuple(
@@ -341,6 +342,8 @@ def run(
 
         if isinstance(val, EventGenerate) or val is None:
             return 0.0
+        elif isinstance(val, datetime.timedelta):
+            return val.total_seconds()
         elif val is Ellipsis:
             if on_exit is not None:
                 on_exit()