According to Media Create yes. According to Famitsu and VGC no.

Famitsu data is through Dec 26th, you're looking at an extra week for VGC.

Is that 190,000 sales for MGS4 greatest hits version included in the sales for the regular MGS4 on this site?

Famitsu and Dengeki have DS ahead, Media create has PSP ahead.  Famitsu has a week different coverage period, but even if they used MC's calendar, they'd still have DS ahead for the year (just by a narrower margin).

VGC had PSP ahead, but has since adjusted it to DS I think?

Yes, VGC includes all budget reprint sales with the original SKUs.  NPD does the same actually, only the Japanese trackers seem to separate everything.

It's worth noting these MGS4 also include some sales from the holiday hardware bundle with Castlevania LoS (probably around 30k or so).

Yes.  In comparison Famitsu has it at 896,434 plus whatever miniscule amount the original version might have sold this year.

For what it's worth, I've heard Famitsu actually covers about 70% of the market directly (which is actually higher than NPD in the US even).  Whatever the case, I'm pretty sure than in terms of direct coverage it's Famitsu > Media Create > Dengeki.

Pro Evo over-tracked then. Think we had it in the 420k or so region.

There is something really screwy with the Japanese sales of Winning Eleven 2011.  Assuming the actual weekly numbers displayed on the Japanese charts are correct, then it is at 392,797 after 9 weeks.

I think I figured out the problem, but not really the cause.  There seems to be a second set of data that is being combined with WE2011.  For the weekly charts

Week 1: 229,503  4,460 = 233,963
Week 2: 233,963 55,189 229,503 1,549 - 4,460 = 515,744
Week 3: 515,744 28,319 55,189 880 - 1,549 = 598,583

and so on (I assume, didn't want to keep going through them).  It seems this other set of numbers is getting added to both weekly and LTD totals seperately each week.  After the first week, the first data set (i.e. 229,503 - 4,460 [225,043 = first set | 4,460 = second set | let's just call the combined number the third set]) of the previous week gets added to the total a second time.

In other words

Week 1: D3.1 D2.1 = W1
Week 2: W1 D3.2 D2.2 D1.1 = W2
Week 3: W2 D3.3 D2.3 D1.2 = W3

D3.1 Would be the third data set, week 1
D2.2 Would be the second data set, week 2
D1.2 Would be the first data set, week 2
W3 Would be the displayed total for week 3

Edit: Ugh, forgot that pluses still disappear after posting.  Well if you don't see an equals or a minus, then it should probably be a plus between any two figures.

So its undertracked by about 70,0000?