Now that is a better looking DSV! Star = good, snowflake = bad!
Like you mentioned, MDX is not for the faint of heart. This was a big stumbling point for me when I first got into SSAS about 5 years ago. When I first attempted to build a cube, I had a Wrox book ("SQL Server Analysis Services 2005 with MDX"), a very complicated relational database that was not structured for dimensional use at all, and no direction for what I wanted my cube to do.
I started walking through the book, came to the realization that my data would not work in the cube, tried to shoehorn it in there anyway, got a ton of attribute key errors (imagine that :-P) and finally got a tiny schema built with one fact and 2 or 3 dimensions. No hierarchies though as I could not figure out how to resolve the infamous duplicate attribute key error, no calculated measures, and no idea what I would do with it. I then got into the MDX chapter of the book and it all fell apart for me. The authors of that book dove way too deep into MDX from the start so that I was discouraged from really learning the language. I then abandoned the project for other work and it took me two years to get back into SSAS. I'm still a bit weak on MDX as most OLAP tools let you get away with a lot through dragging and dropping. MDX doesn't stick with me since I spend most of my time writing T-SQL and have that language burned into my mind.
Good luck with your future cube endeavors. It's not the easiest path to travel but it definitely has its rewards.