Ha! I just saw this, and my first thought was: you stole my idea, LMAO. Seriously, I've been thinking about such a visual designer for Rainmeter as well, but some of jsmorley's points are spot on, and they crossed my mind as well, so I quickly abandoned such a monumental task.
Thing is, this, while a great idea, won't ever be "complete" in terms of features compared to a manual skin design. There are a lot of visual programming tools (e.g. Visual Studio, etc.) nowadays, but even they can't replace the need to manually write your own code - which defeats the whole purpose of a (complete) visual design tool. What this could be useful at would be to speed up and help with positioning stuff in your skin, what happens when you click on an element, stuff that most visual designer tools are able to do. The code needed for something to happen, now that's a different story, and it has to be done manually, to take advantage of the full array of Rainmeter capabilities.
Another thing that would be smart to do, in order to both make the designer adaptable and to simplify your work a lot, is to make it highly modular (i.e. "script-able"). In other words, you just need to create the platform, and let users to add modules/scripts for standard/basic things like weather measures, certain shapes, certain plugin usage scenarios, etc. That being said...Rainmeter is already such a tool (that you can add on small bits of code), so you'd basically do a Rainmeter... for Rainmeter, if this makes sense.
My advice is not necessarily to give up, but start with the basics: positioning, mouse events, resizing, etc. - what a Windows form would do. Create just the base, and design it with modularity in mind, so that other users can add on features in time, in the form of "add-ons". Don't try to make everything - this is next to impossible. Even Rainmeter started just with a few basic stuff in the beginning, and added and refined more over the years...