There are unavoidable problems with a poll like this. Some value though, see comment 3 below.
1. Beyerdynamic, Sennheiser and Audez'e flagships are all present. But the Audez'e's no 2 and no 3 ranked phones are included too, splitting the vote
2. Put another way, a wider net has been cast for Audez'e than the others. It's slightly interesting Audez'e as a brand has the same number of votes today as the T1
3. Variables such as DAC, amp, source material and listening preferences and attributes are uncontrolled. If we assume it is not always possible to pair the 'right' gear with each headphone, then the 'winner' will be the phone that pairs most inoffensively with gear of varying quality. ATM this seems to be the T1
4. Other variables are cost (how easy is it to buy - obviously this has varied with recent economic change), cost proximity (if one can afford a T1 at about $1300, one might step up to an HD800 at $1500 whereas an LCD3 is still out of reach), how long each has been in the market, and how much "band-wagon" momentum developed and for how long.
The significance of points 3 and 4 above is that to compare the number or % of votes, the figures first need to be 'standardized'. However, developing a good model to do this isn't simple - e.g. cost proximity, how time of release to the market interacted with discretionary funds, time of release in relation to competing headphones and so on.
An alternative way of judging satisfaction is the proportion of owners who have kept each phone. Luckily, head-fi provides the figures we need. That said, an uncontrolled factor here is "how long have they owned it?". I'll ignore this for the moment.
Although not every head-fier registers every piece of equipment they own, I'll assume no one brand-owner is more or less likely to register their phone than owners of another brand. This being so, and based on today's own it (O) versus had it (H) figures, a crude index of satisfaction (expressed as a percentage, hence x 100) is:
satisfaction = 100 x O / (O+H)
|Model||Own it||Own it + Had it||Satisfaction (%)|
|LCD2 (combined r2 and not differentiated)||314||351||89.5|
This is imperfect for being just as subject to uncontrolled factors as the poll. On the plus side, the overall sample is larger; n=815 compared to today's poll count of 224. This sample has a much better margin of error than the poll. These margins range from + or - 5.3% (the LCD2) to 7.8% (T1).
These margins mean - for example - we can be 95% confident the LCD2's 'true' satisfaction figure lies somewhere (we have no way of knowing where!) between 84.2 and 94.8%. Notice then that the HD800, LCD2 and LCD3 cannot be separated. Indeed, the T1's true 'I still own it' satisfaction rating could be as high as 90.2%. The T1 is still in the running for 'best' phone - but see the bit in bold below.
The LCD3 sample is too small statistically for worthwhile comparison: margin of error + or - 18%.
The fact there are more LCD2 owners seems commensurate with it being the cheapest. The T1 has fewer owners than the more expensive HD800. This may be evidence indeed of it being closer to the HD800 in price, so head-fiers made the stretch.
The best summation of the above is that all four (five if you unbundle the LCD2) headphones are equally 'satisfying' - i.e. equally likely to be kept. Neither the poll nor the own-it versus had-it index can tell us which is best or worst.
ps: I have the T1 and LCD2. I have started to enjoy the T1 very much since I paired it with a Decware amp; however, the LCD2 is still 'much better'.