I made another effort on finding the FOV for the ATFLIR for better estimates of the target size, and also confirmations on the sensor resolution. Here's what I found:
TL;DR version: the NAR mode most likely has 0.7° FOV, meaning the bird is only half the size of most estimates.
Some history:
Fred started his career with Hughes Aircraft Company in 1982 with the Electro-Optical Systems group in El Segundo, CA. After establishing the Hughes Thermal Imaging Evaluation Facility, he shifted his focus towards the development of a newly-emerging class of high-performance IR imaging sensors based on staring focal plane arrays. In 1993, Fred's team demonstrated the world's first InSb 480 x 640 format (20 um pitch) MW Imaging sensor. This achievement led to a series of program awards for U.S Navy submarine periscopes and Navy/Marine aircraft sensors. including the AN/ASQ-228 ATFLIR for F/A-18 Hornet and the AN/AAQ-27A for the V-22 Osprey.
SBIR New Product and New Director of Engineering Announcement
Competitor features, more pixels on sensor, but still just 1 megapixel:
The fourth generation version of LITENING, built by Northrop Grumman, features the 1024 x 1024 pixels FLIR sensor for improved target detection and recognition ranges under day/night conditions; new sensors for improved target identification; and other advanced target recognition and identification features. Other product improvements include a new 1k charge-coupled device sensor, which provides improved target detection and recognition ranges under daylight conditions.
http://www.theatlas.org/index.php/about-atlas-3/honoring-dr-john-warfield-5?download=69
ATFLIR specs (it also lists 4 other products with FPA 640x512 and one with 320x240):
name: Raytheon AN/ASQ-228 advanced sight FLIR(ATFLIR)
detector: InSb FPA 640×480 (centerline space 25μm)
waveband: 3.7~5.0μm
field of view: Wide: 6.0°; middle: 2.8°; narrow: 0.7°
Application status and development trend of infrared imaging system
ATFLIR specs again:
Raytheon AN/ASQ-228 ATFLIR, recently named Terminator
Dimensions: L: 183 cm/72 in D: 33 cm/13 in W: 191 kg/ 420 lb
Capabilities: 640 480 FPA FLIR operating in MWIR: WFOV 6 6; MFOV 2.8 2.8; NFOV 0.7 0.7
Carrier aircraft: F/A-18A þ , C/D,E/F Replacement for AN/AAS38 Nite-Hawk
http://www.helitavia.com/books/Mil Av Sys/Wiley - Military Avionics Systems.pdf
There's plenty of general info about IR sensors there, IR tracking and so on, such as:
The FPA provides an image on to a focal plane that coincides with the sensing array, most usually a two-dimensional array whose dimensions scale easily to a standard rectangular display format: NTSC, PAL and, more recently, VGA and XVGA and above, greatly simplifying the optics. Although the figure depicts the focal plane array with a vertical axis, in tactical systems the array face is usually facing directly towards the target. In most cases the forward looking IR (FLIR) sensor is looking forwards, the term being relative as it is usually mounted upon a gimballed assembly that has extreme angular agility and slew rates in order to be able to track targets while the platform is manoeuvring. As will be seen later in the chapter, several EO sensor systems are commonly physically integrated into the co-boresighted sensor set to aid sensor fusion and allow target data to be readily handed off from one sensor type to another. In an array the entire surface is not given over to IR energy sensing. There is a certain overhead involved with interconnecting the array which prevents this from being the case. In a practical array the overhead is represented by a term called the fill factor which describes the useful portion of the array as a percentage. On modern state-of-the-art arrays, the fill factor is usually around 90 %. The array is effectively read in a sequence of frames in the same way as any other real-time imaging device. Therefore, the time between successive read-outs of the array image is the time available for the array to ‘capture’ the image, and this is referred to as the integration time and permits successive images of the target to be generated. The key element in the performance of any IR imaging device lies in the performance of detectors and the read-out of the imaged data in a timely fashion.
...
The typical desired capabilities of a modern sensor are summarised below:
Pixel pitch: ~20–40 μm
Frame rate: 50 Hz (PAL); 60 Hz (NTSC) with a desire to go to 100 Hz and above
Maximum integration Time: 99% of frame time
Data rate: 10 MHz upwards
Array size: 640 x 480 (VGA resolution), heading towards 1000 x 1000 (1 Mpixels) or above in next generations
Articles about Raytheon products and IR sensors in general. No ATFLIR specs but the most interesting part is about another product, AN/AAS-52, giving further indication that the FOV can be different for IR and TV modes:
Fields of View, Degrees
Wide: 33 X 44, Medium-wide: 15 X 20, Medium: 5.7 X 7.6
Narrow: 1.2 X 1.6 (IR&TV)
Ultra-narrow: 0.6 X 0.8 (IR)
Ultra-narrow: 0.22 X 0.29 (TV)
RAYTHEON BRINGS EO TECHNOLOGY To Defend Our Nation - engineering
Some general info how IR tech works:
http://iopscience.iop.org/article/10.1088/1757-899X/51/1/012001/pdf
So, did those help? At least there's more than enough confirmation the sensor resolution is 640x480 (VGA), and the pixel pitch is probably that 25μm (so a lot bigger pixels than what high res full frame DSLRs for example have, meaning the FLIR is able to gather more light per pixel).
But what about the FOV? There were two additional documents stating the narrow FOV is 0.7°, just like that Raytheon marketing material. That 1.5° seems to be based just on the info given in that flight sim page, which I would regard less reliable.
But we need to consider the possibilities that 0.7° meant FOV with 2x (digital?) zoom, or that it's only available on TV/electro-optical mode. That digital zoom explanation doesn't seem to fit based on that one document I quoted above:
The wide FOV is optically fixed at 1X magnification. The medium and narrow FOVs are optically fixed at 1X with a 2X magnification zoom capability.
If 2 out of 3 modes provide 2X zoom, why would they give only one of them with the zoom and the other without? That sim page also claimed all of those modes would have 2x zoom. I don't think it can be trusted.
But even more significantly, consider what these options mean if the middle has 2X zoom capability:
field of view: Wide: 6.0°; middle: 2.8°; narrow: 0.7°
2x middle is 1.4! What use would that have, if the narrow was 1.5? Similarly, it's no surprise wide doesn't have 2x zoom, as it would be just the same as the middle mode. 1.5° narrow just doesn't seem to make any sense.
But then there's still the possibility that those above modes work so only for the electro-optical channel, not for IR, as I also quoted earlier above:
The ATFLIR package was developed to provide the Boeing F/A-18 combat aircraft with an austere weather, long-range targeting capability. It is housed in a 72-inch pod on the port fuselage station. Common optics and a mid-wave staring focal plane array support an infrared channel with 30x magnification and an electro-optical channel offering up to 60x magnification.
Another version
Rainey noted that with the ATFLIR pods, pilots could safely see crisp, clear subjects from nearly 4 miles up and 30 miles away. An infrared targeting system his planes used magnified an image in the night by 30 times. In daylight a television camera can magnify it by 60 times. These capabilities are 3-5x better than previous FLIR systems
Pilots in Theater Clamoring for ATFLIR Pods
I remember reading the optical path was supposedly the same for both, and those documents indicated the modes work by moving mirrors, so why would it be different for IR? It would fit if the IR couldn't use the narrow mode but only middle with 2x zoom, or that it couldn't use digital zoom on the narrow mode, but both the Nimitz and Gimbal videos use IR with NAR and 2x zoom. And what are those magnification figures relative to?
Other sources state "ATFLIR's magnification is 30X versus previous FLIR capabilities at 4x", previous generation being AN/AAS-38 "Nitehawk" with 3º x 3º and 12º x 12º modes, fitting to that 4x difference. But where does that 30x (let alone that 60x) come from?
AN/AAQ-14 LANTIRN PODs in F-16s had 21ºx28º and 4ºx4º FOVs, and that was also replaced by the ATFLIR. So if they compared a FOV of 21º to 0.7º, that's exactly 30x difference, and with the 2x zoom it would become 60x.
Edit: I may have found where that 30x could come from. From that Raytheon marketing doc:
Built-in navigation FLIR (optional)
From that book:
A relatively wide FOV for navigation and target acquisition (~20º x 20º)
So maybe the ATFLIR has a 21x21 optional navigation FLIR, and that 30x is the ratio between that and the narrowest optical FOV of 0.7ºx0.7º.
Then there's the question of why didn't those pilots try that 2x zoom on the "Go Fast" video to see more detail, even though they used that in the same mode on "Gimbal"? Did some restriction prevent them from using it while contrast tracking or looking downwards or something, or they just didn't bother to try to actually see what that blob was?