2009 Subaru Forester 2.5l NA SOHC Rough idle, P0302 missfire

I'm new here so apologies if I'm not going about this the right way.

I just did a full rebuild and all the ignition components are new, at first I couldn't get it to run at all, turned out I set the valve lash incorrectly, then I had an issue with my crank position sensor(bent pin), after fixing those two issues I still have a noticeable misfire at idle which reports cylinder 2, P0302.

Compression is on the low side of spec for all cylinders except #2, just a hare under spec, I should have bored over but I just want 5 more years out of it. I haven't performed a leak down yet as I don't have the tool handy. But my research leads me to believe it's an ignition side problem.

I'm having trouble interpreting the secondary waveform, I can only get one cylinder at a time because it's hard to get at the Subaru primary and I only have one clamp at the moment, I've included pics from cylinder #2, but I get similar results on other cylinders, just not as bad.

The overview is only one cylinder, compression, and waste. Periodically you can see the compression firing voltage is high, the spark voltage is high and stays super high(this happens as often as once every two compression strokes, and as seldom as something like once every 10)

Looking at the detail, you can see, a decent burn line, sometimes it looks like it doesn't loose any energy after the burn, sometimes it looks like it looses a bit but after the ring out, it stays quite high. It looks like an open short, but the coil, cables and plugs are all new, and I can't find any evidence of carbon tracking.

drive.google.com/open?id=15fru3YWn2mZJRh_2GSNZ5ghdUqyWfHAp
drive.google.com/open?id=1jBQPTDevVRE1Jscsk2CRbIW6wUHr2r2p

I'm just hoping something jumps out at somebody or a suggestion for the next tests I should perform. Is it possible I'm chasing some artifact and I should look at some of the other waves, not just the ones that stick out like a sore thumb? Should I get an amp clamp and dig into the loom looking for that primary? I haven't investigated the fuel side at all yet because that high firing and spark voltage before the midpoint has me looking at the ignition side.

Thanks in advance.

UPDATE

I noticed that it runs a LOT better in open loop so I took a closer look at my MAF

drive.google.com/open?id=0B4r3d0Z-x1hKVX...Y2xEeGdZMlVKY2g2cm1r

it's reading below 2 at idle, Subaru spec is 2.5-2.7. Basic diagnostic is, check the power, with engine off, was just about exactly battery 12.2, speck is greater than 10, now I have to test the cable resistance back to the big plug just to make sure, but looks like I need a new MAF, I'll update after my next test.

Stop me now if you think I'm going down the wrong road, that's a pricey little bugger.

What were the compression numbers? I'd be inclined to verify mechanical integrity before moving forward.
Timing marks still spot on? Not doubting your aptitude but I've seen seasoned veterans get it wrong. It happens.
If compression is low then the engine may not be breathing well so the the MAF may be accurate. You can try unplugging it and see if it runs better, but that's not exactly fool proof.
The other thing that comes to mind concerning an idle misfire on a single cylinder that improves with throttle is a valve train issue.
I haven't had the pleasure of doing an 09, but from what I recall of the older ones the engine layout isn't exactly convenient for valve adjustment.

I can't recall the cyl pressure at the moment, 3 were in spec but low, and one was a couple PSI low, they were all fairly close to each other. But the test may be invalid, I didn't have the throttle body open which I'm guessing could cause it to report low, I'm going to do a leakdown test tonight, hopefully that will give me some confidence in the mechanical.

No offence taken re my aptitude, I'm an amateur, don't assume I did anything right. I triple checked the timing last time I had it open, with the MT guide in place I don't think it could skip, it's all new components and it's fairly taught.

drive.google.com/open?id=1EWto5yXeHwbcqMLa39DExFWJN6WkkMBS
drive.google.com/open?id=1EX_g06h24F0z4PsLj-ZRFkY8VDIwlJXA
drive.google.com/open?id=1EXvuGfYZUuBg15ePu5Il-3xj9GMy1em0

The right bank is blurry, but the lined up skinny paint mark has the notch and it's dead on, the center timing marks are dead on, the replacement oil pump (from Subaru mind you) doesn't have the mark at the crank sensor like the original unfortunately so it's harder to see, but I lined it up with the case split.

That left bank cam mark makes me nervous, but if I moved it over to the left one tooth, the line was no where near the mark, my options were just to right as you see it, or way to the left or way to the right. I gave it MANY complete rotations, the pulleys always lined back up every 2nd rotation and rotation was smooth (except for the compression resistance) and no contact was felt.

When I said it ran a lot better in open loop, I got it to go into open loop by unplugging the MAF, the engine smoothed right out, rpm was more stable, waveforms looked better, not perfect, but better.

One of the reasons I think it's breathing ok(maybe not great, but ok) is the combination of the MAF reading low and the fuel trim being maxed out, this is my reasoning, the MAF reads low so the the computed fuel is low, the O2 will read a lean condition and order positive trim. If the MAF was good, and it just wasn't getting much air, then it shouldn't be reading really lean and pinning the trim.

At idle MAF is a good 30% low, if it was a breathing problem, would going open loop improve it that much?

You're right about the valve room, I just pulled the motor last time I did a valve adjustment.I find the valve lash tricky so it could still be off a bit. Lets just say I totally messed it up the first time, luckily the only consequence is that the lash was too big, that mistake can't cause interference. The second time I made sure I was on the compression stroke by feeling the air in the spark plug for each cylinder, so I know I was at the top of the compression stroke, but was I TDC perfect... it's hard to line up the mark, however I know I'm pretty close, so it might not be perfect but it should be really close.

I found a cheap MAF on amazon, I'm going to throw it in tonight and see how it reads and runs after I do the leak down. If that reads as expected and it runs better, I can re-evaluate from there. If that does turn out to be a problem I'll order a quality part, I just couldn't get a quality one that quickly.

Thanks for the suggestions.

I think you got the marks as good as you're going to. Let us know how the MAF works out

The new MAF didn't change things at all, reading was the same, I still have to double check the harness, but I suspect it's fine. Good thing I found a cheep one. The chances that they are both bad and reading low roughly the same amount is pretty unlikely I think.

The leak-down was kind of all over the place, I want to re-do it being more careful, they all reported in the green, but it's a new tool so the green doesn't really mean anything, worse, they had a span of about 20% (input was about 100psi, steady was 70-90psi maintained, or 10-30 leak if you prefer) most of the leaking air seemed to be through the rings, which still aren't fully broken in. No leaking through the coolant. I didn't detect any intake or exhaust leaking but I didn't look too hard as it was easy to hear the leak through the oil fill tube. Next time I'm going to drop the exhaust anyway so that should make it easier to check for leaks there. Interestingly, the cylinder with the highest loss was not the one with the lowest compression and it's also not the one that is reporting the miss.

I don't the leak-down explains why I'm pulling in so little air, it's 30% low at idle.

The vehicle was run for quite a while in a bad state, I think I'll check the caddies and make sure the outflow isn't clogged up.

Any other suggestions are welcome.

After a cold winter in the garage and early spring yard work, I finally got back to the problem yesterday and I did two more tests, one of which was very compelling. First I put an amp clamp on my injectors to make sure they were firing, I already tested them outside the engine, I just wanted to verify that they were getting a good signal, recall that I'm reading low air on the intake side, but super lean on the exhaust side* so I wanted to make sure it wasn't a fuel problem (already tested the fuel pressure too) The second test was an intake leak test with propane (I already did a vacuum test and was getting a good vacuum) and sure enough, the engine smoothed right out when I gassed the drivers side intake area.
Here it is: Last time I pulled the engine to fix the valve lash issue, I failed to torque down the driver side intake bolts... The first time I did it, I marked each bolt as I torqued them, but the second time... they were already marked. Now that I've torqued them down, engine smoothed out, intake air flow is reading on spec, O2s are looking good... I might have my Subi back. I'll try to remember to post a follow up after a couple months of testing.