Attention IPVMU Grads! FOV Calculation Verification Requested

Slogging my way thru the IPVM back-catalog I came across this post about viewing a face at 3000 meters.

I might have clicked out had John's following statement not caught my eye (emphasis added):

Wow, 3000m is really really really hard.

Bombast and hyperbole you say? Though a frequent user of the singleton 'really', he has (to my knowledge) only ever employed the doubled 'really' once, and never til now has been a triple-threat.

However, before I risk the ridicule of the lurking quants in the original post, I would like to run this proposed solutiion by anyone who feels they can validate it, IPVMU alumnus or not...

Here's the proposed camera: (Cohu HD 8800)

Required Distance: 3000 m

Focal length: 1680 mm

H Sensor Size: 9.398 mm (1/2.7")

HFOV = Distance / (Focal Length / H Sesnsor Size)

16.75m (55 ft) = 3000m / (1680mm / 9.398mm)

PPF = H Resolution / HFOV

35 PPF = 1920px / 55ft

35 Pixels per foot is at the very bottom range for face recognition, but can anyone say that 35 PPF is even correct?

Thanks!

P.S. For extra credit take a look at this graphic included by John in his reply to the original poster:

Since 1/3=.333 and 1/2 = .5 and 2/3 = .666 and 1/1.8 = .555 wouldn't you expect to see the focal length of the 1/1.8 one between the 1/2 and 2/3 ones, yet it is smaller the the 1/2" one....

What am I missing?