I usually eat all of my fitbit cals from my ChargeHR, if not more, and I am still losing. I have MFP set to sedentary, lose 1 lb a week, and plan to exercise 5 times a week for 30 mins. Fitbit is set to the same I think but I don't log my food there either. The ChargeHR factors in your heart rate to come up with a calories burned I think. Are you sure you will gain or are you just afraid you will? If you are just afraid that you will gain, you could experiment for a couple months with eating most of them back and see what happens.

daily calorie burn is step based.

So I'd confirm that it's calculation of your stride length is correct since you have new knees and perhaps may not stride out the same as average - or maybe aren't anyway. Like longer/shorter legs than average your height.

How did you determine your stride length?
It should be based on average daily pace - not the extremes of the range like grocery store shuffle or exercise level walk. But the middle, and actually better to be more in what occurs the most.
Like just a serious daily walk pace on a known distance track - longer the better. Across a room is not good.
Start an activity record on Fitbit so it shows the distance for that. And try to hold to that non-exercise pace.
I'd also use that as opportunity to count maybe 100 right foot steps and confirm the Fitbit number went up by 200.

Also, the adjustment is based on what MFP thought you'd burn with your BMR and activity level, with no exercise.
Not your eating level.
Fitbit daily burn - MFP estimated daily burn = adjustment.
Eating goal + adjustment is actual eating goal (I know you got that part).

So Fitbit 3201 - 1207 adj = 1994 MFP daily burn.

1994 - 1240 eating goal = 754 deficit - which math rounding probably selected 1.5 lbs weekly.

1990 / 1.25 sedentary = 1592 Mifflin BMR being used. Might confirm that in MFP Tools for BMR calc using your stats.

So rest of the math sounds right (if BMR is confirmed) - it's merely the amount of adjustment based on distance and pace - which is all based on stride length setting and accuracy of steps seen.

Do you commonly carry anything of decent weight on daily walking at say work that would increase weight?
This could increase impact on steps leading Fitbit to think you could be jogging. Higher calorie burn there too.
Of course it should go up if that is case, just not that much.