"One servant has both first and last name, which suggests a free person of color who has been hired as a servant. Most of these officers do not appear to have owned slaves, so several other servants could have been hired from their owners. Complexion of two servants is given as white, so these two officers must have hired free white men or boys."

I will have to take exception to two of your points in the above quote. Concerning last names, some slaves did have last names. This was most common among slaves coming from larger plantations - the use of a last name was helpful in differentiating between several Charlies, or Johns, or other such common names as might be present among the slaves. One plantation in Louisiana gave slave the last names of their previous owners. Other plantations might have used other methods for giving last names, but a last name is not necessarily an indication of non-slave status.

Concerning complexions - a "white" complexion is just that, a complexion. It does not necessarily indicate race. Many quadroons, octoroons, or people of color of lesser black blood than that may well have had very light ("white") complexions and Caucasian physical characteristics as well, and been able to pass for white/Caucasian where they were not personally known. Slaves with white complexions were not out of the question. Take a look at the picture of some of the Reconstruction politicians in Louisiana after the War, and you will see what I mean.