I agree Purrin.  I have a hard time believing it's precisely tuned garbage. 

What I've discovered tuning orthos follows what Purrin said, and is why I made the list explaining the steps you go through when designing a headphone.  The signature most headphones have is the easiest one to design and have a resolving/clear sound.  Since they've gotten away with it, and since their primary goal is profits, I don't think they've seen any reason to put more effort and money into refining their products more. 

It would be so easy to make a T50rp sound super open, airy, clear and detailed but with peaky response.  I managed to do that right away and if I was Sennheiser, Beyer, Grado, Ultrasone I would have had a finished headphone a long time ago with much less effort, or in their case R&D investment.  As I outlined, the first thing you get is clarity.  It's really challenging to get clarity without peaky, resonant response you have to be really precise. 

so I think it's laziness and cheapness.  Look at the T5p.  There's no way that response is on purpose.  It's just the first thing they got that met what they were looking for in clarity/resolution. 

As an aside, I just wanted to express how irksome it is seeing head-fiers post "flat tuning is boring."

Glaringly obvious v-shaped turning is evil, and evil is banal. In hell the devil walks in circles. 

I think people like to think that it's not a treble boost but rather...the headphone have more clarity and are more revealing.

I have a friend who boosts his hi-fi system to something like +15 @ 8kHz and +18 @ 10kHz and he uses this to play vinyl. 

He hears it as clarity and just loves it for that. When I EQ'd it for him a bit, he felt his system lacked resolution, speed, dynamics, and clarity. He also felt it sound less natural. I have been honest with him and told him his system is one the most bright systems I have ever heard and that he just likes it bright. He just smiles.

But mainly, because it's CHEAP to produce. Membranes will naturally vibrate (ring) at a certain frequency. Think of dynamic transducers are drums.  "Faster" material which is stiffer, tighter, etc. will ring at higher and more specific frequencies. Lots of engineering effort needed to damp those resonances.

This, basically.

Companies don't follow a perfectionist method the likes of RD's 4-prong approach and little care is taken when integrating drivers. Then there's the lack of human testing due to the expense of setting up a proper trial, though companies like AKG have every opportunity to use HK resources. For a minor headphone company just starting out I can see most of the testing being outsourced to another firm, maybe even a marketing firm. The only feedback becomes conflicting end-user response ("treble spikes of death!!" v. "most clarity evar!") and so the loop falls apart with no improvement to the design.

It takes work to tame the inherent non-linear/sometimes-random vibrational tendencies of a transducer. Smaller companies often buy drivers from someone else and stick them in cups that only look good enough to sell. Or beyer, who design their own glamorously magnetic drivers and stick them in the same decades-old shell they use for nearly the entire high-end lineup. Even when a company takes the time to consider both the driver and the shell, like Sennheiser did with the HD800, you might still get a peak as a compromise for widened bandwidth and lower non-linear distortion. Senn then went ahead and combined the smaller company and beyer ideal by modifying the magnet structure of a stock driver and sticking it in a version of the HD800 shell, a compromise that added a couple more peaks.

Peaky treble exists because physics is a bitch and companies fail to follow through on proper R&D.