If your body had adapted by slowing down - 2 weeks is not enough time, though it might have started to go up. If you were with a suppressed metabolism, then the excess was just truly surplus for a while and you added fat, until such time body felt like speeding up.

In one recent study it took 3 months eating at maintenance to regain 1/2 of what had been suppressed. In other words, daily burn was 500 less than it needed to be, after 3 months it was still 250 lower. But that was also eating at lab measured maintenance, not slightly above to encourage faster increase to metabolism.

So no, that 2 weeks was mainly fat added, and hopefully an increase, so keep it up.

Since you have very variable sessions for exercise, you need to eat more on the days, mostly afterwards, that have the longer runs.
So you can't do a average TDEE method, or you'd be so far low on calories on big run days, and so excess on low run/rest days.

Here's a modified method that works well.
Only count 1 hr of any runs that you do for deciding TDEE level, and your lifting time. So that may be 5 hrs on top of otherwise sedentary 45 hr deskjob/commute weekly. If you have kids or pets and lawn care and long shopping sessions, ect, then you aren't sedentary, and you need to go up a level just for that aspect. So Very Active.
If you are sedentary otherwise, and I could see recovery needing you less active outside exercise, then select Moderately Active.

Now, on running days where you go longer than 1 hr and you really need the extra calories. Estimate your calorie burn for the extra time over 1 hr.
During reset eat that all back. When you start a deficit take the same 15% off and eat that amount back.

That way you can get in the extra carbs and calories you need after a long run so you can do it again tomorrow, but you still get to average out some of the workouts over the week so you have some idea of eating level to plan on.

What may also help here, not only in getting TDEE, but also on doing that latter method.
Spreadsheet on my profile page.
Stay on Simple Setup tab and get stats filled in after looking and deleting the sample data.
When you get down to the Activity Calc, do just what I said above, lifting workout time, and only 1 hr of running time on planned days.
Now go over to MFP Tweak tab.
Put in the Katch BMR shown, and the daily activity multiplier shown.
There's your TDEE with only 1 hr of the runs accounted for. That's the reset goal, don't use the deficit yet. But that deficit is based on amount left to lose from Simple Setup tab, so it may not be 15% if little left.
Down below is where you can put in the extra time above 1 hr for any run and the calorie burn for that extra time, and it'll remove the deficit and already accounted for calories, and give amount to log and eat back. Again, that's for when you start your deficit.

How do you calculate your running calories now?
Do you have something that logs the distance and time and elevation to get a good estimate?
Because a 50 cal underestimate isn't bad with 30 min jog, but 3 hrs that's really bad and would impact performance from eating too little.

Yes, use your brain, not your stomach. You know you just burned alot. Are you aware of the research regarding increasing carb uptake for muscle restore within 30 min after endurance cardio?
I'm hoping you keep up with stuff if doing this level of cardio. Usually once you get some food in ya, or drink, the correct hunger will come back.

And you might look at all these numbers to convince yourself extra bonus of calorie burn isn't a wise idea.

Like if you burn normally no exercise say 2000 in a day, and you just did a long workout that burned 2000 too - that's massive to what you normally burn.
Same as what you burn in 24 hrs, in 2 hrs.
And then if you eat 1500 say - wow. Just gave your body less than half of what it burned. You can bet a body isn't going to like that.

You probably have no inkling of the performance you could have with a really healthy body, being fed well enough.
People seem to appreciate vitamin and mineral deficiencies, either by getting some ugly side effects when it happened and then watching their diet better, or being pre-emptive and taking daily pill.
Calorie deficiency is just as bad.

You might compare your HRM to this to see how accuracy is. Several of the models can estimate VO2 nicely under athletic level, but once it goes up they start falling very short.
http://www.myfitnesspal.com/topics/show/774337-how-to-test-hrm-for-how-accurate-calorie-burn-is

My simple 1 hr run yesterday of 7 miles was best estimate of 904 calories, Garmin, even with measured HRmax stat, only gave 731.
I'd expect performance to suffer if I was taking a known deficit and that was adding on another almost 200 cal deficit. And imagine how awful on my 4 hr ride Sun to go by the Garmin. Ugh.
In fact that did happen spring of last year, and lost decent LBM, before I got another VO2max test. I'd been using data from prior one 1.5 yrs early, and formula for calorie burn based on it was way underestimating by that point.