So I've tested those cells. My testing methodology is what is generally accepted in the E-Bike and EV communities. Which is to use this inexpensive LiPro Balance Charger / Discharger:
http://www.ebay.com/itm/Imax-B6-LCD-Screen-RC-Lipo-NiMh-Li-io-n-LiFe-Battery-LiPro-Balance-Charger/350782413357?_trksid=p2045573.c100505.m3226&_trkparms=aid=555014&algo=PL.DEFAULT&ao=1&asc=44941&meid=75c77cf0df84417d8ddc675e954db692&pid=100505&rk=1&rkt=1&
First I charge all the cells at 1C. In the case of unknown C rates I usually charge at .5C of what is printed on the cells. So for these Panasonic 3400mah my charge rate was 1.7amps. I smart charged each cell individually to max capacity (4.2v). Then I discharge at 1amp to empty (3.0v). Record the measured mah and then charge at the newly found C rate. Then I check the measured mah going back into the battery to verify the margin of error.
This is a video of a guy testing random batteries to show you how the tool works:
So these Panasonic 3400mah NCR18650B MH12210 Made in Japan batteries are as follows:
Cell 1 - Discharge mah = 1998, Charge mah = 1961
Cell 2 - Discharge mah = 1998, Charge mah = 1959
Cell 3 - Discharge mah = 1998, Charge mah = 1958
Cell 4 - Discharge mah = 1998, Charge mah = 1967
My sample size was small (4 cells) but I'd say they consistently test 1998mah. These are holding 59% of what they claim. At $5.75 per cell I'd say this is a bad value. You could argue the legitimacy of the batteries I bought or the quality of the LiPro tool but I urge everybody on this forum to at least question the "rated" capacity of the cells they are buying.
There are some LG that are rated at 3000mah that people show tests anywhere between 2800mah-2950mah. I'd consider looking into those before these Panasonics even though they are in this BOM thing. I will track some of these down and test them as well. I will probably start a thread in the Battery section to record a variety of the cells I've tested.