To me, this is a question of whether you want to capture the set of knowledge about measureable body sites, or only the question of whether some particular body site in a condition is measurable (as determined by some method not otherwise represented in the FHIR resources).
In either approach, the Measure resource is not so much about whether or not things are measureable, but what is the formula for calculating the measure (i.e. median time from admit decision to admission, or percentage of patients that have been appropriately screened for HIV).
In the simplest case, where you just want to indicate whether a body site is measureable and you’ve made that determination elsewhere, I’d recommend an extension on Condition.bodySite that indicates whether that particular bodySite is measureable.
If it’s representing knowledge about measureability of bodySites in general, I’d recommend using the BodyStructure resource with an extension about whether that site is measureable.
Alternatively, you could use a terminology to represent both the knowledge, and the application of that knowledge to a particular Condition.bodySite.