Here is the deck at 9:14. The sun had been creeping up to that table from the railing for the last 30 minutes; I had put the Solar Pathfinder right in front of the table yesterday to do the measurement.
The location I measured got sun around 9:25, after which the sun moved gradually moved away from the able as it got blocked by a different part of the roof. At 10, it looked like this:
So the solar window predicted by the Solar Pathfinder seems pretty accurate (which means I used it right, I don't doubt it is an accurate instrument).