Comparing DNA results with First Cousin

My ethnic breakdown autosomal DNA 

My First Cousin's Ethnic breakdown

My first cousin Susan DNA tested with 23andme and her results just came in. Susan shares my maternal line. When I saw the results I immediately noticed that she did not have any Middle Eastern? Her mother being my mother's sister I thought this was strange. My mother's results included 18 percent Middle Eastern. My results showed 13 percent Middle Eastern. My mother and I tested with Family Tree DNA. My mother's Middle Eastern was said to be Mozabite. I also thought it was odd that no Spanish showed up in Susan's results? Odd considering her grandmother Forgey was from Nicaragua?
Some of this inconsistent information probably has to do with the different testing procedures at the two companies. Also, ethnicity predictions are subject to debate. It's possible to determine the continent of origin of our ancestors using DNA testing, but pinpointing the exact country of origin is generally not possible. Isolated populations are probably much easier to make a connection with. Since Susan would have inherited half of her autosomal DNA from her mother it's possible that she didn't inherit the Middle Eastern associated markers? It's also possible that Aunt Dorothy didn't inherit as many Middle Eastern markers.
My cousin and I share approximately the same amount of Native American DNA, so this result seems quite accurate. Family Tree DNA doesn't show trace amounts of DNA such as the 1.4 percent Sub Saharan African Susan got from the 23andme test. I did find those traces when I used the Gedmatch admixtures and Doug McDonald also found a small amount. After taking the MtDNA test I discovered my maternal MtDNA Haplo group was Sub Saharan African. My cousins MtDNA is also Sub Saharan L2a, so we have confirmed that is correct. She actually got a more specific Haplo group of L2a1. I had heard that haplo L2a1c was Northern African but after doing more research it seems to be more widespread. So our Sub Saharan African is either from a slave from central Africa or a North African?  I had assumed that this result was connected with the Mozabite but it could be unrelated. Since MtDNA mutates very slowly and can remain unchanged for thousands of years it can be tricky to make assumptions based on it. Half of those who matched me on the HRV1 portion of the MtDNA test have African roots, and are descendants of former slaves, and the other half had Semitic roots. When I uploaded my HRV1 and HRV2 results to the Sorenson DNA site I found an exact match from Mali. This would tend to point to our L2a being from an African slave. It's probably impossible to say for certain though. Sometimes these tests raise more questions than they answer.

Sorenson MtDNA match