With a car that was now fairly leak free and comfortable to drive I started putting some miles on it.. This is where my big troubles began..Part of me must be a descendant from Niki Lauda.. God blessed me with an ok mind, but a really great ass.. I feel things going on in vehicles most people don't.. For example, I had no issues jumping into a 2005 S2000 at autocross and drive it at the limit, with VTEC kicking in attempting to send the car for a spin cycle, mid corner.. But jump into a 2006 when they switched to DBW, and all I did was spin out due to the lack of response in the DBW system.. If you've ever been blessed with racing an S2000, you'll know the dangers(or should I say fun?) of mid throttle VTEC..
So what does that mean for me? Well little issues, 90 - 95% of the population wouldn't notice drive me NUTS!! My drive to work now included 2 rather large hills.. flat land the car was good, but tip in on the throttle to maintain speed when climbing the hill and I would get a light hesitation/misfire... Before the advice pours in, don't worry I've got this. Sit back and enjoy the ride... my specialty for 10 years was diagnosing problems like this..
Lets start with basics...
Check your dry flow air filters... They wear out over time.. This car came with a Corksport TIP and Power series 2 intake on it, but it was still on the factory tune..

Something I learned a long time ago... Air flow restrictions on the intake side of turbo car have a major effect on spool up and how well the turbo works... The 2nd Mazda 323GTX I owned was setup with a bunch of bolt ons and chipped ECU.. I had my manual boost controller set to 10 pounds of boost.. One day I upgraded the OEM intercooler, to a bigger 626 intercooler, upgraded the piping and built a new intake pipe up to the VAF (old school MAF sensor for those of you who don't know).. First test drive with out touching my boost controller, the car hit 18lbs of boost...
I don't like K&N filters/oil filters.. So why the *beep* would I used one?? Well locally it was all I could get with out having to wait for a custom order.. I wanted something larger... I didn't pre-plan replacing the filter, so I took what I could get that day...
Part of my testing path, was to have scan data.. I was hoping I'd be able to make a post with evidence around cleaning your MAF sensor". When I saw how bad the stock filter was.. Cleaning your MAF sensor in order to "fix" a problem is wrong... It doesn't work.. When a MAF sensor gets dirty enough to cause a problem, it's developed a baked on crust that skews it's calibration.. This will show up as slightly negative fuel trims at idle, positive fuel trims under load, that gets more positive the more load you put the car under. Plus lean with knock issues during open loop/WOT conditions.. You can "improve" the maf sensor enough to get a check engine light to turn off and reduce these issues by cleaning it... That's why it's sooooooo heavily recommended...
So why isn't this a good "fix"... Well you can't totally get the crust off the sensor.. The crust slows the sensors response time down.. Scan data isn't fast enough to catch what goes on during quick throttle transmissions.. You need a lab scope to catch this... With the Mazdaspeed platform, you wont hear it... But on other vehicles, you'll get "pinging" during throttle transitions. Despite how dirty my MAF sensor was, it wasn't causing problems..I've got some of my pictures tucked away on a hard drive somewhere, I'll come back and update this post if I find them..
Cleaning a MAF sensor in order to maintain it and prevent the crust from building up, is ok.. It's not a bad thing to do.. Generally when you do this, you wont see any changes in your car. If you do see a change, if it does make the problem better, or go away.. Replace the sensor...
A tip in, hesitation/misfire (Something to understand is different types of misfire and what they feel like, mine was "soft" subtle.. plugs,coils, and some of the quick "go-to" advice isn't the common problem in this case.).. A MAF sensor issue that's got a bunch of built up crust on it as I've talked about is a likely culprit.. It's also the easiest and quickest to test...