How do you calculate calories burned?

I think you could reasonably log the stretching side of the intensity scale as "Stretching, hatha yoga", the strength training as "Strength training (weight lifting, weight training)", the cardio as "Calisthenics, home, light/moderate effort".

This takes a little explaining, especially that last one. Behind the scenes, MFP uses something called "METS" to estimate exercise calories. The METS for an exercise are generally determined in research studies. Very generally, METS are a number that can be used in conjunction with our current body weight and the number of minutes we did that thing to estimate the calories burned. One could kind of think of it as a "relative intensity" of the exercise.

There's a big database of METS values on the web for different exercises. The only METS values I can find for chair exercises are in the older adults section of the Compendium. (I don't know whether you're an older adult or not, but the arithmetic difference isn't huge, so it's probably close even if you're younger. No estimate is perfect, but all we need is to be workably close.) That says most chair exercises are in the range of 2.0-3.5 METS.

The MFP cardiovascular exercise database entries I suggested in the first paragraph are approximately in that same range. I looked for ones that relate to what you're actually doing, in the description. The only weird one I'm suggesting is that calisthenics one for cardio. That's about the closest description I could find in the right METS range. Most of the other options (aerobics and that sort of thing) assume more intensity, presumably because of the leg involvement when it isn't chair exercise, which adds calories.

I hope that makes sense. No matter what the exercise is named, a (say) 3.0 METS exercise is going to yield the same calorie estimate in this methodology, for a person of X bodyweight doing it for Y minutes, regardless of what they're doing in a exercise-name sense.

You could create custom exercises for these chair exercises using the nearest METS values and a METS calculator outside MFP, but that would be more complicated. The above would be the best simple-ish idea I can come up with for estimates that ought to be in the ballpark IMO.

FYI, the compendium I mentioned is here:

https://pacompendium.com/

And there's a generic METS calcultor here (one of various ones on the web):

https://ergo.human.cornell.edu/MetsCaloriesCalculator/MetsCaloriesCalculator.htm

If you use MFP's cardiovascular database, it does the METS calculator part for you, behind the scenes, using its METS values for the exercises listed.