You can see in the discussions on those issues that there is some consensus that the content of Darwin Core terms should remain consistent with the definitions and not mix concepts.

The definition of the sex term (without looking for clues in the Comments) might be seen as a little ambiguous, "The sex of the biological individual(s) represented in the Occurrence. Recommended best practice is to use a controlled vocabulary."

Strictly speaking, there is no single sex that matches the biological individuals in this example lot Occurrence record, and so the sex fields should be left blank.

There are a number of terms in Darwin Core whose definitions suggest that a list (formatted with ' | ' as a separator) be used to match multiple values. Though it might seem that sex could be one of these terms, it isn't currently. Again this argues for a single value for sex from a controlled vocabulary.

The biggest problem is with the Comment on the sex term, which gives an example with different semantics from what the definition says - namely "8 males, 4 females". I pose that this is an error and must be corrected.

Yet, all hope of retaining information in the Occurrence record is not lost. To capture the richness of the information in the multi-sex lot example, I would recommend the use of dynamicProperties. 

How should one populate Darwin Core terms for a record of 2 males and 5 females in a single lot? I would do this:

sex = null

individualCount = 7

dynamicProperties= { "count of males":2, "count of females":5 }

Hope that makes sense.