A good sanity check is to compare your results with what the New Horizons team is reporting. It is also useful to consider whether there are any other factors at play that could throw the numbers off.
As a point of reference, this undated article on the New Horizons website discusses the Ultima Thule flyby and states:
The pixel sizes of the best expected color and gray scale images and infrared spectra will be 330 meters, 140 meters, and 1.8 kilometers respectively. There is a chance of higher resolution grayscale images, with 33 meter pixels, if the high-resolution LORRI camera is able to point accurately at Ultima, but the necessary accuracy isn't guaranteed
Calculated resolutions
For reference, the estimated diameter of Ultima Thule is about 30 km.
(The estimated size of Ultima Thule is discussed in more detail in this answer)
Using the calculator you linked to with a distance of 3500 km I get the following:
MVIC (color and greyscale)
Sensor width: 5000 pixels
FOV: 5.7°
Quoted maximum resolution: 330 m/pixel
Theoretical best resolution: 69.60 m/pixel
FOV at target: 348.30 km
LEISA (IR)
Sensor width: 256 pixels
FOV: 0.9°
Quoted maximum resolution: 1800 m/pixel
Theoretical best resolution: 214.65 m/pixel
FOV at target: 54.95 km
LORRI (greyscale)
Sensor width: 1024 pixels
FOV: 0.29°
Quoted maximum resolution: 140 m/pixel (high probability of imaging UT)
Quoted maximum resolution: 33 m/pixel (low probability of imaging UT)
Theoretical best resolution: 17.29 m/pixel
FOV at target: 17.71 km
Limiting factors
The calculated numbers clearly don't match those published by the New Horizons team. I see three reasons why this could happen.
Blur
Both MVIC and LEISA are pushbroom scanners and because of the low light levels at Ultima Thule they will be slewing the platform slower than they did at Pluto. This is less of an issue for LORRI since it doesn't use the pushbroom technique but it too is affected by the low light levels and thus also needs longer exposures. This could lead to increased image blur particularly since the target's position is imprecisely known (this may result in incorrect pan or slew rates). Due to their design, MVIC and LEISA are particularly affected.
Field of view considerations
Ultima Thule is a small target with an uncertain position. To improve the odds of imaging the target the field of view needs to be kept large; this reduces the ground resolution. To further improve the odds and to allow higher resolution images to be taken, multiple images will be taken in a grid pattern covering the area where Ultima Thule is likely to appear. This should allow (fairly) high ground resolution images to be taken but at the expense of taking up precious time and having a lot of blank frames. Keep in mind that capturing any image at all takes priority over a chance of taking a detailed image.
In the case of LORRI there's an addition consideration: the imager's restricted field of view. At closest approach LORRI only has a field of view of ~18 km which is considerably smaller than Ultima Thule's estimated 30 km size. Time constraints (see below) mean that at closest approach it is impractical to do extensive grid scanning with LORRI; this decreases the odds of a non-blank image being returned. I calculated that the closest LORRI images that are expected to show Ultima Thule will be taken at a range of 14000 km (high imaging probability) and 6500 km (low imaging probability).
Speed and distance
New Horizons will be passing close (3500 km) to a small target (and considerably closer than it did Pluto). The relative speed is also high (14 km per second). The window of opportunity for imaging will be much smaller than at Pluto and slew rates and parallax shift while imaging will be higher; this will exacerbate the two aforementioned issues.
Ultima Thule imaging schedule
Source: NASA (cropped, edges cleaned up)
This is the imaging schedule as published in the press kit.
Here's a rough breakdown of this diagram with distances given in km from point of closest approach (not ranges) accurate to within about 100 km:
LEISA (IR) -65750 to -64000, -34900 to -26000
MVIC (color) -42500, -17400 to -15400, +25600 to +26850
MVIC (greyscale) -10500 to +2250, +5600 to +10300, +22200 to +23400
LORRI (greyscale) -61800 to -60000, -42400 to -42000, -31900 to -23100, -17000 to -15100, -10250 to +2500, +2800 to +4400, +5900 to +10500
Conclusions
The images taken during the encounter are to be acquired from a variety of distances under less than ideal conditions. Image quality will therefore be lower than calculations based on a stationary observer at closest approach would suggest.
Resolution during capture in meters per pixel
Resolutions shown assume perfect image capture (i.e. no blur).
Time SCET Distance Range LEISA (IR) MVIC (color) (grey) LORRI (grey)
.------.
-4553 04:17 -65750 65850 | 4040 |
-4432 04:19 -64000 64100 | 3933 |
'------' .-----.
-4280 04:22 -61800 61900 | 306 |
-4155 04:24 -60000 60100 | 297 |
'-----'
.-----.
-2943 04:44 -42500 42650 | 849 |
'-----' .-----.
-2936 04:44 -42400 42550 | 210 |
-2909 04:45 -42000 42150 | 208 |
.------. '-----'
-2417 04:53 -34900 35100 | 2152 | .-----.
-2209 04:56 -31900 32100 | | | 159 |
-1801 05:03 -26000 26250 | 1610 | | |
-1600 05:06 -23100 23350 '------' | 115 |
.-----. '-----'
-1205 05:13 -17400 17750 | 353 |
| | .-----.
-1177 05:13 -17000 17350 | | | 86 .
-1066 05:15 -15400 15800 | 314 | | |
-1046 05:16 -15100 15500 '-----' | 77 |
.-----. '-----'
-727 05:21 -10500 11050 | 220 |
| | .-----.
-710 05:21 -10250 10850 | | | 54 |
0 05:33 0 3500 | 70 | | 17 |
156 05:36 2250 4150 | 83 | | |
173 05:36 2500 4300 '-----' | 21 |
'-----'
.-----.
194 05:36 2800 4500 | 22 |
305 05:38 4400 5600 | 28 |
.-----. '-----'
388 05:39 5600 6600 | 132 |
| | .-----.
409 05:40 5900 6850 | | | 34 |
713 05:45 10300 10900 | 217 | | |
727 05:45 10500 11050 '-----' | 55 |
.-----. '-----'
1537 05:59 22200 22450 | 448 |
1620 06:00 23400 23650 | 471 |
.-----. '-----'
1773 06:03 25600 25850 | 515 |
1859 06:04 26850 27100 | 539 |
'-----'
Values derived from distance data shown in the Imaging schedule diagram.
- Time Seconds relative to closest approach
Derived from distance assuming a velocity of 14.44 km per second. Negative for approach, positive for departure.
- SCET Spacecraft Event Time on 2019-01-01
Derived from time based on closest approach of 05:33:30 SCET. Format is hh:mm.
- Distance Distance in km to point of closest approach
Estimated from diagram. Negative for approach, positive for departure. Accuracy is about ±50 km.
- Range Distance in km from spacecraft to 2014 MU69
Derived from distance assuming a closest approach of 3500 km. Rounded.
Initial images
The images planned to be sent during the initial (daily) pre- and post-encounter transmissions are:
MVIC (color) 900 m/px (NYT 2)
LEISA (IR) 28 wavelengths at 1800 m/px (NYT 3)
LORRI (greyscale) 8000 m/px (Failsafe 1), 5000 m/px (Failsafe 2), 300 m/px (NYT 1), 140 m/px (NYT 2, low probability), 140 m/px (NYT 3, different image, low probability)
These will presumably be lossy-compressed images with lossless ones being transmitted later.
Sources
I hope that's helpful. I'll go over it again in a bit and see if I can improve it.
