You got a box of apples and oranges there, can't really compare.

FitBit is estimating non-moving calorie burn on calculated BMR from gender, age, height, weight.
You selected Scooby for Katch BMR with weight and BF%.

FitBit is estimating moving calorie burn on weight and perceived step based movement which can be very accurate.
You selected from 1 of 5 rough levels on TDEE calc.

FitBit is estimating in exercise but you removed that to get a non-exercise TDEE figure.
You selected a TDEE level that probably included exercise, right.

So you can't compare either BMR or TDEE and get anything meaningful out of them. But you can use the differences to see if it might be accurate.

You say you have your BF% saved on FitBit. As a stat that is merely tracked, or actually used? I've not seem them change their method of calculating BMR, because majority have no BF% to use. But perhaps they do, might check the FAQ's as to what they base their BMR on.

What is your Katch BMR?
What is your Mifflin BMR (last FitBit FAQ said it uses that or close enough)?

That difference is what matters for all non-moving time. May be meaningful, may not be.

Your moving time is based on weight and step based activity, is your exercise that type, or possible bad things for it to estimate, like their FAQ says, rowing, lifting, cycling, ect?

Neither of those is going to help knowing how your thyroid effect will be on weight loss, starting with best estimate and watching the results will. Which for a woman means a month test at a time sadly. Because your BMR literally does change through the month.

Curious how a nutritionist could test metabolism, unless they did the RMR test with gas exchange capture, which is only valid method you'll find.

Also, if taking meds to regulate thyroid, metabolism is probably close enough to expected, and you can accidentally force it lower by eating too little.

Oh, HRM will over-estimate calorie burn on lifting. The formula for estimating calorie burn from HR is totally based on aerobic exercise, amount of needed oxygen for burning fuel, and steady state where HR isn't changing much for 2-5 minutes. Lifting is anaerobic, without oxygen so heart being has nothing to do with that, and opposite of steady state as HR goes up and down.
Just use MFP's estimate of calorie burn for strength training, it may seem low compared to cardio or HRM, but it's true.
Also, the aerobic exercise level starts at about 90 bpm, so if skating doesn't get that high, it'll be inflated there too.

For better way of using FitBit with this program of Deficit % rather than block that MFP uses, use this spreadsheet.
Fill in Simple Setup tab up to but not including Activity Calculator (unless curious).
Now go to the FitBit_BodyMedia tab and fill in known info there.

http://www.myfitnesspal.com/topics/show/961054-spreadsheet-for-bodyfat-bmr-tdee-progress-tracker

So Cunningham is an RMR calculation, and RMR should be 150-250 over BMR, and indeed it is for your Katch BMR, also based on bodyfat%.

And Mifflin inflated isn't that awful bad, 125, with TDEE that would be almost 200 inflated at Moderate level, and with 20% deficit, only 155 inflated over best deficit.
And since FitBit is using something similar for non-moving time, that means only 1/3 to 1/2 of your time daily, so 75 cals, not really enough to worry about, though it is over 500 cal's weekly inflated.

Just remember that when you add exercise and log it, log the calories minus same %.
The spreadsheet MFP Tweak tab will do that automatically for you in lower section.

How do you know HRM is right on for lifting calories?
Find your avg HR for some lifting session, and then do a treadmill session and nail that same avg HR, and see what the calorie burn is. Depending on HRM, it'll be the same. But in reality the lifting was less calorie burn, as most of that avg HR was anaerobic elevation and recovery back down, stress, and pressure, nothing to do with burning calories.
Now if Garmin using Firstbeat algorithms, then much better estimate.

Definitely a good plan starting on the high side, especially before an RMR test.

So you mean the FitBit gets to see no standard non-exercise days? You could do Method 1 with almost non-exercise days, almost doing total eatback, but including some of your days as spread out.

FitBit does some assumptions if not worn and not seeing some activity. I don't think that works, just replacing the values with your own works.

But you could make a workout start stop of normal activity of sitting, see how many cal / min that averages out too, and then replace your exercise with that value to get a non-exercise TDEE.

Some sources say Cunningham is BMR, but it's really RMR formula.

Inflated values, higher than reality, or at least higher than better estimates.
Let's say best case scenario, you got your RMR tested and were not undereating, and got 1400, meaning the BMR would be about 1250 say.
You got a Bodpod too and indeed have less LBM and more FM (Fat Mass) than expected for your age, height, weight, and the 1250 is actually expected for that LBM (that would be Katch BMR).

So then you do a Mifflin BMR calc to get TDEE, only based on gender, age, weight, height, and get estimated Mifflin BMR of 1500.

So that calculated value is obviously inflated over reality, by 250 calories.

Then you pick your TDEE level of Moderately Active 1.55 x BMR and say that is indeed correct, and get TDEE 2325. That TDEE is now 388 inflated over reality.
You take a 15% deficit and eat at 1976, but that is actually 330 inflated over reality.

Now, if you had used your test figures, BMR 1250, TDEE 1938.
You'll see that eating to inflated value of 1976, even though a deficit with the math, actually has you eating over your real TDEE of 1938.

So someone in this case during reset, eating 388 over real TDEE, would continue to gain weight right up until they went to a cut, and at cut would actually be at maintenance.
So while lifting they would indeed slowly trade LBM and FM, but they could be do more for fat and weight loss, they could be taking a safe deficit off real TDEE and be eating 1647, and actually lose fat and weight at the same time.

This happens more than many on here realize, because if you've done VLCD or even bad yo-yo dieting for years, you've progressively eaten away at your muscle mass, and now have less than average and expected by the non-Katch BMR formulas. Less muscle also means less glucose stores. Less metabolism, ect. It may actually be appropriate and expected for their amount of LBM, just a lot less than other formula, I've seen many times 200-400 inflated BMR and TDEE values, and no actual deficit in place compared to best estimates.

For HRM, if HIIT is done correctly (all out anaerobic push followed by recovery to do it again (sounds like lifting, huh?)) then it's not right either. HRM thinks you are hitting that high level as aerobic, in which case the heart is beating that fast to provide required oxygen to burn the fuel.
Fact is your primary and if done right only fuel source is carbs, which doesn't even require oxygen to burn, so the amount your heart is beating isn't related at all to oxygen delivery required.

I tested a Polar with correct tested stats (I think yours has VO2max also) to a Garmin which uses algorithms appropriate for non-aerobic stuff (it watches breathing, so if breathing doesn't match HR, it's anaerobic, and not counted).
Lifting with 8-12 reps, 1-2 min rests between sets, was about 1/3 to 1/4 what Polar estimated.
HIIT with 15 sec on, 45 sec off, was about 1/2 to 2/3 what Polar estimated.
And confirmed HRM's working right by doing aerobic activity at same average HR as that entire workout, Polar had about the same calorie burn (which was appropriate for cardio) and Garmin had much higher values matching Polar closely.

Not sure what that means to have calorie estimate turned off on FitBit.