There are no potential performance hits. In particular the code generated for list comprehensions would be faster than your example because there is no extra function call per item.