You don't have to re-run the ACBC analysis (don't have to re-compute the utilities). The way I can think of doing this is creating new segmentation variables (with the names you prefer) that are copies of the segmentation variables that were given to you.
This is a bit clunky (or I can't think of a slicker approach right now), but it should work. Unfortunately, it requires setting up a bunch of "filters" in the filter manager; then creating a new segmentation variable that relies on those filters.
Go to Analysis + Filters and add a bunch of new filters that point to your cryptic variables. For example, let's imagine you have a cryptic variable called "abcd" that takes on values 1, 2, or 3. You can create three new filters called (for simplicity) a1, a2, and a3. The logic for a1 is abcd=1; the logic for a2 is abcd=2, etc.
Now that you've set up three filters that point to the three values of that original cryptic variable name, you next create your friendly name for that segment (that you'll include in the simulator file that you build for your client). Go to Analysis + Segments & Weights and click "Add Segment". Call this segment your friendly name. For the categories of this segment, indicate that category one uses filter "a1", category two uses filter "a2", etc.
Now you've got your friendly segmentation variable names that are copies of your original cryptic variable names. When you create your simulator package file for the standalone simulator, you'll tell it to include those new friendly segmentation variables.