Total time of 48:05 for a little over 6.5 miles.
TLDR: I ran 4 x 1 min hard/2 min recovery during second half of the run.
Full version: I found an app that works with my watch that makes the GPS actually function to a respectable degree, so I decided to try it out on an actual run. I tried the app out previously walking a 0.54 mile loop and it read 0.53 miles at the end, with the discrepancy due to slow pickup of GPS at the beginning, so I was hopeful that for a 6+ mile run the amount of error would be reasonable. I made comparisons by running a 6.03 mile loop that I had previously measured on the course tool. My general belief is that my course tool measurements are almost never short, and if they are long the measurement is likely not off by more than .01-.02.
At the beginning of the run things didn't start off so well, as the delay in GPS recording led to the watch showing 0.02 miles run after only 0.05 miles. However, after the start things began working much better as at the mile mark the discrepancy was still only 0.03 miles. Shortly before the 19 minute mark of the run I had to stop at a bathroom, and my measurement was messed up as I had to run off course. I paused the watch, and upon resuming there was once again a delay in which the watch did not record any distance run for a few seconds. Based on extra course tool measurements I will estimate that the added distance due to the pit stop was approximately 0.06 miles, of which the watch only recorded 0.04.
As I was running up the hill on the backside of Alpine Blvd, I ran 4 x 1 min hard, with a 2 min recovery. I ran pretty hard for the first two reps, and with a less severe uphill grade on the last two reps I felt like my motivation to push myself was not as strong as it should have been and I did not really start pushing myself hard until the last 15-20 seconds or so.
With a mile to go in the run my watch read somewhere between 5.01-5.02 miles (I was in the middle of my last 1 minute rep, so I had a hard time catching the distance at the mile marker), and at the finish my watch read 6.01 miles. With the added distance for the pit stop I am estimating that the actual distance for the loop was 6.09 miles, giving me a total watch-run discrepancy of 0.08 miles. Of this distance, 0.05 can be accounted for simply based on GPS lag upon starting/unpausing the watch, indicating that the mid run error was only 0.03 miles. I feel like this amount of error is reasonable considering the error associated with my own measurements using the course tool, and it is substantially better than the standard 2-5% error that I would experience when using previous software on this same watch (0.08 mile error is only 1.3% assuming that the course tool measurements are accurate). What is additionally positive is that I would expect the amount of error to be smaller as my runs get longer (as long as I do not stop frequently) considering that over half of the error for the total run was related to GPS acquisition errors when starting the watch.