Well the major languages I looked at (C, Java, Perl) just ignore leap seconds, as does the POSIX standard. If you ignore leap seconds, you're not UTC and saying you are is incorrect. Maybe you're actually just TAI, but probably not since the language APIs don't know about SI seconds and work on the assumption that there are 86400 seconds in a day. But since it's a linear timescale, I can at least convert to and from another one when doing astronomical calculations.
I haven't checked but I suspect the situation is as bad on the Microsoft side of things, given that those guys still completely fail at DST adjustments on a regular basis. It's difficult to imagine them getting the occasional extra second every now and again correct. And really this situation would be fine except that everything always seems to break whenever anyone actually tries to use NTP to handle it correctly.
So if I have a hypothetical database of satellite locations that are a month old and the spec calls for them to be stored in UTC, it immediately becomes impossible to correctly plot their locations on a map. Especially since the spec for the inputs also says UTC but the identifier in the file always indicates "Zulu Time". Which I believe is just GMT. So you already have a 26 or so second probable error putting the coordinate into the database (~15 miles off the satellite's location) which will only get magnified if anyone else between you and the database tries to do leap second accounting again. This is kind of a problem if you're trying to hit a target within half a meter on the surface of the Earth with a "Laser." You'll be aiming at Saddam and accidentally hit a French embassy. Entirely hypothetical example.
Of course, if you start needing sub-second precision (Say, for targeting a femtosecond "laser" at the surface of the earth,) you might need to start thinking about relativity, since your atomic clock on your satellite will drift from your atomic clock for your ground station at sea level, even though both of them only ever lose a second about once every few trillion years. And they'll both be correct. Then you start to realize that the universe is just some poorly-written n-dimensional graduate student's thesis project, demonstrating how to convert hydrogen into plutonium, and retire to a profession where you don't have to deal with any span of time shorter than a season.