The problem is compounded by the fact that a total lack of knowledge about materials and engineering has never been a bar to every cyclist feeling entitled to an opinion about what constitutes a 'comfortable bike' and what makes it so.
Most people focus on the construction material alone. But the vast majority of bikes are made up of tubes, which are then arranged into structures. The properties of the raw material is only one of a myriad of variables that will effect the characteristics of these frames. When you then add in; wheel construction, seat pin length, saddle rail length, handle bar rigidity, tape thickness, tyre build, size and inflation pressures, you have even more reasons why a very subjective view of 'comfort' will be so varied.
But to say that carbon fibre is not capable of building a comfortable bike is just plain wrong. There are many different types of carbon for a start which can all be built into structures with an infinite range of characteristics. I incorporate carbon fibre into the most flexible structures (fabrics) and the most rigid structures (highly loaded cantilevers). No other material can meet that range of applications.
It's no coincidence that the part of a bike most under load, where the greatest benefit can be gained from getting the correct balance of strength, flexibility and comfort and where above all, reliability is paramount (ie the forks), has been the first part to switch whole heartedly to carbon.
And beyond the forks, there is in fact no application where carbon couldn't also build the best solution from an engineering point of view. You could have the lightest or stiffest or strongest bikes ever and some combinations of them, plus you could also have the most comfortable. A material that allows the construction of fly fishing rods, archery bows, wind turbine blades and helicopter rotors could easily cope with a bit of bike flex. In fact in different configurations, carbon is essential in the more radical 'comfort bikes' that are built when designers don't care about the UCI constraints.
But bikes have to also meet market demand and retail price points. So far, it's been easiest to sell carbon on its lightness and 'pro levels of stiffness'. But you only have to compare the Cervelo RS and R3, or Spesh Roubaix and Tarmac, to see how you can make two similar bikes ride so utterly different.
The same debate used to rage about 853 and 531 tubing. Everyone knew that 853 bikes were stiffer, but the two different steels had absolutely the same degree of elasticity. However as 853 was stronger, it allows thinner walled tubes to be used, which saved enough weight to use tubes of a bigger diameter without a weight penalty. And bigger diameter tubes are much, much stiffer. This was compounded by these tubes were used predominately to build race bikes where 18 or 20c tyres were the rage. The ride was bone shakenly hard, and so everyone soon knew that 853 was so rigid you couldn't build a comfortable bike out of it. But if you used the extra strength to only build thinner walled tubes of the same diameter and then made up a similarly dimensioned bike with identical equipment as the 531 tourer, then the 853 bike would be more , not less comfortable.