Estimate TOTs for packages.

We estimate the longest possible time from mission start to TOT for
all flights in a package and use that to set the TOT (plus any delay
used to stagger flights). This both cuts down on loiter time for
shorter flights and ensures that long flights will make it to the
target in time.

This is also used to compute the start time for the AI, so the
explicit delay option is no longer needed.

gen/briefinggen.py

@@ -1,3 +1,4 @@
+import datetime
 import os
 from collections import defaultdict
 from dataclasses import dataclass
@@ -116,9 +117,10 @@ class BriefingGenerator(MissionInfoGenerator):
         assert not flight.client_units
         aircraft = flight.aircraft_type
         flight_unit_name = db.unit_type_name(aircraft)
+        delay = datetime.timedelta(seconds=flight.departure_delay)
         self.description += (
             f"{flight.flight_type.name} {flight_unit_name} x {flight.size}, "
-            f"departing in {flight.departure_delay} minutes\n"
+            f"departing in {delay}\n"
         )
 
     def generate(self):