In high school, I learned calculus without the help of basic math skills, and instead persisted on sheer force of will and my ingrained ability to follow explicit instructions. Similarly, I have learned to bake–nothing pleases me more than a set of instructions without compromise, without leeway, without the intrusion of the human soul. I will found the Hermetic School of Baking and finally know peace.
Because of this, nothing in baking confounds me more than red velvet cake. The red does not contribute to the taste at all (the smooth depth of the cake comes from cocoa powder countered with vinegar and a little more salt than usual). There is no Vulcan logic reason why red is more successful a velvet than any other color. Yet, according to totally objective experiments with color, red has some kind of arcane magic that truly makes it the king of velvet-based cakes.
I used a recipe which had previously yielded 24 cupcakes, and divided it evenly into thirds. I made each third identical, save for the dye color. Red yielded the predicted 8 cupcakes. Green was reduced to 6. And blue had a measly 4 cupcakes. What the hell, blue. Pictorial evidence will show that the blue cupcakes even look denser. They formed a kind of smooth crust, even though the entire cupcake top would squish a little on command.
I call for red to yield its dark secrets to the rest of the baking world. I want my deep, rich crayon colors to be equally delicious across all desserts.
I refer you to the history of Red Velvet Cake as recorded by Wikipedia: http://en.wikipedia.org/wiki/Red_velvet_cake#History
Maybe it was the dye you were using? And/or the different chemicals that make the different colors. I wonder if the results would change w/the use of more professional grade dyes.