My vote was both...
I want a visual representation in the same IDE I am programming in so I can visually architect what I am programmatically making (I don't care that it isn't a word).
Also, if a visual tool negates the need for twenty command line entries, then putting your nerd tendencies away in favor of efficiency would be the correct choice. If you were to build a script that could be dropped in and configured to your liking using your code, would you still shun it? If not, perhaps it should be inspected whether the tools being offered implement best practices. If so, then you'd be a fool to duplicate the work already done for you. Perhaps these tools do not give you the full range of features you need, but as long as you can fall back to the underlying code and tweak to your needs, you may have saved yourself a few dozen lines of code. Done several times, the time savings is immeasurable. No, it should not be done in lieu of understanding the underlying code, and that is not what I am saying.
Ask if a designer was given the option to code his way through a design, would he? Of course not, and many of us developers need to remember that we are many times called to be GUI designers, regardless of platform. With that in mind, the sterility of code does not give one the creative representation of ones work. You should always be vigilant of your aesthetics, as this is ultimately what your end user will see.
I know I'm arguing two points (and one slightly off topic), but I am specifically on board with what Microsoft Visual tools has done, and believe that adopting those principles would assist even the best of developers if handled correctly.
