Well, time has now passed and we can see whose test prediction
for southern California was the winner - all in fun, of course.

The predictions were:

KATE:
Ml 3.2-4.8 within 35 km of 34.4N 117.72W between 6/28/01 (5:45PM)PST
and 7/05/01 (5:45PM)PST

LOWELL (FIRST PREDICTION)
I think the earthquake will occur about 80 km SW of your stated epicenter
several days after your prediction expires.

MARK:
Ml 3.2-4.8 within 80 km of 34.4N 117.72W July 4, 5 or 6

LOWELL (SECOND PREDICTION)
Ml 2.5-4.0 within 35 km (later revised in jest to 200 km) of 33.9N 118.6
on July 7.

Twelve well-located mainshocks of Ml>=2.5 occurred in southern California
during the period covered by these predictions (from SCSN catalogs)
The listing of these events follows:

6 30 2 50 59 2001 34.88N 116.41W 0 2.6 1
6 30 14 21 20 2001 34.29N 116.77W 4 2.6 2
6 30 22 30 54 2001 35.80N 118.05W 8 2.7 3
7 1 16 57 8 2001 32.78N 116.16W 4 2.7 4
7 2 8 44 19 2001 34.00N 117.43W 8 2.5 5
7 2 15 25 26 2001 33.25N 117.48W 12 2.8 6
7 3 2 0 20 2001 35.93N 117.68W 3 2.5 7
7 3 11 40 48 2001 34.26N 116.76W 3 3.9 8
7 3 11 43 13 2001 34.26N 116.76W 3 3.0 9
7 04 22:27:53 2001 33.20N 115.60W 3 2.5 10
7 05 23:31:18 2001 34.28N 116.48W 2 2.6 11
7 07 02:28:36 2001 33.53N 117.92W 16 2.8 12

How to evaluate these predictions depends on what you consider most
important in a prediction. If the event is required to be within the
magnitude range stated, than for Mark's and Kate's predictions only
one event would qualify - the Big Bear event of Ml 3.9 on July 3 (# 8).
All would qualify under Lowell's criteria.
If location is considered the most important criteria, then
distance to the predicted epicenter is the most important. None
of these events fall within 35 km of Kate's or Lowell's predicted epicenter
and only one occurs within Mark's 80 km radius. 10 of the 12 occur
within 200 km of Lowell's predicted epicenter.
On the other hand, if timing is considered the prime target,
11 of the 12 occur within Kates 7-day window, 2 occur within
Mark's 3-day window and 1 occurs on the day Lowell expected.
Comparing predictions is obviously a no-win situation. so let's
try making all three parameters equally important and producing
some kind of point system.
Starting with any event in the expected window with a
magnitude in the expected window would seem a good starting
place. In this case, Kate would have one event - # 8, there
are no events of Ml>3.2 on July 4, 5 or 6 so Mark's prediction
is out based on this arbitrary criteria.
The Big Bear quake was 90 km from Kate's predicted epicenter,
but in the correct time range and magnitude range.
But the Newport Beach earthquake of July 7 fell within
Lowell's time range and magnitude range as well and occurred
only 75 km from the predicted epicenter.
It would seem that the most accurate prediction was the first
made by Lowell in which it was stated that the event would occur
about 80 km southwest of Kate's epicenter several days after the
end of her time window (i.e. 7 or 8 July). The Newport Beach
earthquake occurred 90 km southwest of Kate's epicenter on
July 7.
I leave it to the reader, however, to decide for him/herself
which of these predictions best matched the available
facts. This is a hairy problem not for the squeamish.

I would love to see how other's deal with this comparative
problem - any takers?