Let's do some calculation.
angle of Murray Street = 120.87°

some relations:

a = 118px
b = 339px
green angle = solar azimuth - Murray St angle
For all possible sun position that casts a shadow that reaches down to the sidewalk but don't cast a shadow on the sidewalk we can say:
SIN(green angle) = a/d
SIN(solar azimuth - Murray St angle) = a/d
ARCSIN(a/d) + Murray St angle = solar Azimuth
ARCTAN(b/d) = solar elevation
d is all we miss but we can say that
d min = a
For all possible d>a we get combinations of solar azimuths and elevations:

Now we just need to find out which combination fits any possible combination on 11th Sep. 2001!

The orange curve represents the way of the sun on 9/11 between 9:00am and 12:00am.
The blue curve represents all possible combinations of azimuth and elevation for a shadow that reaches down to the sidewalk but don't cast a shadow on the sidewalk.
The intersection of both curves is at 10:18am (but without Daylight saving time!!!).

Finally we should estimate the possible failure of the result.
May be the image from the video is somehow distorted or the perspective distorts the relation of a/b.
It's about 1px per minute for the length of a. In other words, if a would really be a=108px then we would get about 10:08am and for a=128px we would get 10:28am.
The green marquise has a special feature: a round corner. The viewing angle doesn't matter! We always see the right relation between the hight of the wall and the width/radius of the green marquise.
The Google Earth image shows the green marquise just 101px wide. For this relation we get a time between 10:00am-10:10am.

A later time than about 10:20am would require a much wider marquise than measured in both images but since the green squares below the marquise don't show any visible distortion a later time can be excluded.
There is another way to confirm the time. The video pans to the opposite side of the street. For a short moment a lens flare effect becomes visible that points directly at the position of the sun.

This frame...

...shows the facade. The left edge of the frame is approximately aligned with the corner of the building. The roof line is visible too.
The corner of the roof corresponds to a solar azimuth of about 150° and therefore a time of 10:40am. Furthermore we see that the vertical lines are bowed.

Obviously the video was shot using a wide angle optic that caused some pillow distortion. That distortion stretches the horizontal axis and shortens the vertical axis. Therefore the relation of wall height and the width of the marquise should also be distorted in that way.

If so then we get an earlier time than 10:18am.

The lens flare effect points at an azimuth of about 140° (about 10:10am).

Nevertheless, there is a major mistake in all these times because Manhattan had Daylight Saving Time.
That means it is 1 hour later already and about
11:10am EST.

What time is it?

It seems it's 11:11am.

