Let's be clear about one thing right from the start: there's no such thing as a perfect road bike. What makes a comfortable bike for me might be impossible to ride for you. What feels overly stiff to me might have exactly the kind of responsiveness you're looking for. And that paint job over there: eww! I wouldn't ride that in the middle of the night.
Well, you get the idea. Where there are no objective standards there cannot be an objective evaluation. While this is perfectly fine if your sole purpose in life is to ride your bike as hard as you can, this might be a bit of a problem if you're trying to sell bikes to someone else. Or if you like to compare your stuff to that of others.
To resolve this issue, we might want to introduce some objective standards: mass, stiffness and aerodynamic drag forces, to name but a few. While we can argue for hours about the color of your bar tape, there's no point in discussing the stiffness-to-weight ratio of your new bike: it is what it is, whether I like it or not—the beauty of science in a nutshell.
However, the problem with this is what we did when we introduced our objective standards: it wasn't science, it was engineering. While science is the thing that tries to find out how things work, engineering is the thing that tries to make things that work better. You could say that science concerns itself with understanding the universe, and engineering concerns itself with optimizing it.
Having a bachelor's in mechanical engineering and a master's in spacecraft engineering, I have optimized a few things in my life, from roller bearings to deployable sails for satellites. And having 4+ years of experience as a Ph.D. student in mechanical engineering, I have seen the engineering approach applied to all kinds of things, from dinner parties to holiday planning. Most engineers simply like to optimize stuff, and that's okay.
The basis of all optimization problems is understanding the relationship between the parameters you can fiddle around with (your inputs) and the properties of what you're trying to optimize (your outputs). For example, increasing the diameter of your bicycle wheel is a great way to reduce rolling resistance. Inputs. Outputs. Optimized!
Problems that are easy to optimize are usually based on the laws of physics: larger wheel means less elastic deformation means less rolling losses. Stiffness, weight and aerodynamic drag forces belong to the same category; each one of them can be optimized with a little (or a lot) of calculation effort. The result is a bike that's very stiff, very light and very aero. However, the laws of physics don't say that any of the above is the same as "very good"—they're just straightforward optimization problems compared to the somewhat nebulous goal of making a "better" bike.
What makes mass-market bicycle design so difficult is that there is no straightforward relationship between inputs and outputs if your output is "perfect bicycle"—and it gets worse if your output is "perfect bicycle for any given person". If you optimize something with everyone in mind you will end up with something that literally fits nobody—ask the US army.
So what is the solution then? Do we need to optimize each bike with a particular person in mind? Do we all need to ride bespoke, custom-made road bikes? Or is there a middle ground? Are there any objective parameters that we can optimize so that our bikes are designed for everybody? What materials should we use? What color should we choose? And how will we decide if last year's model is better than this year's if we deem shedding 25 g of frame mass irrelevant?
The short and obvious answer is the universal mantra of all scientific endeavors: it depends. And this blog is dedicated to finding out what it depends on.
Unfortunately, we won't come across many easy answers on the way, except for one: Do we really need to be engineers to build good—or if you so will: perfect—road bikes? The answer is what I call the paradox of bicycle design: You don't have to be an engineer to build a perfect road bike—but you might need to be one to realize that.