I would be concerned about correctness, and authority.
The Creative Commons licenses make a lot of sense for spatial data that have key "creative" components - an index of delicious restaurants, or some other subjective topic. The attributes in such a dataset bring more value than the specific location, and a restaurant woudn't require within-a-meter accuracy.
However, for a dataset of gas pipelines, I want that to be as correct as possible! Granted, defining "who may use this" does not deliver a guarantee of correctness, but the nature of an open license makes it harder to place a value on the data.
Consider a published dataset of pipelines, licensed with the Creative Commons license. Someone later makes an update to that dataset to reflect what they feel is the true location of a certain pipe in the ground, perhaps due to construction or erosion or some other event. Now there is a disparity. Is this a technical concern, or a subjective / interpretive concern?
Update: A quick Google search exposed this document:
SPATIAL DATA LICENSE AGREEMENT
(I use this as an example)
Under Section V, Use Restrictions and Other Prohibited Activities:
The Licensee is familiar and understands the provisions of the National Map Accuracy Standards and assumes the responsibility and liability for the use of this data at other than the compilation scale (scale at which the digital data was intended to be output in hard copy format).
So, this license does not assume responsibility for the accuracy, but it does transfer responsibility to the licensee clearly. The Creative Commons license assume that the copyright holder has a 'creation', not a 'reference', which is the distinction I'm trying to make. I imagine one could add a clause to your Creative Commons license about accuracy.
I feel that Open Source software is a bad analogy for Spatial Data under a Creative Commons license. Open Source allows the freedom of any user to inspect the insides and validate it for their own satisfaction. Open data does too! Except... I don't want to have to validate it. Validating the position of something is not the same as validating the correctness of some lines of code; if I have to valid something's position on the Earth, why bother using someone else's?