Which just proves that frequency response curves mean sweet FA in a real world.
On axis - anechoic - what has this to do with reality.
Not sweet FA, but they need viewing with caution.
Anechoic responses do give you a flavour and will also indicate any serious problems.
They need to be viewed in combination with room averaged plots which evens out the smaller peaks and dips. These are far less audible than broader trends.
Most listeners tend to listen close to on axis, if not directly on.
That's why a range of plots are useful.
If a speaker has a ruler flat response on axis, but a huge crossover suck-out when you move say 45 deg off axis - you will hear this effect on axis because you also hear the reflected sound in the room, and this of course includes the suck-out. It is one reason why good crossover design is so important and a good crossover has good responses both on and off axis.
So we need more/better measurements, not less!
Just a point on the ES14.
It is meant to have aperiodic loading which when implemented carefully gives most of the benefits of a sealed box but lowers the main system resonance, allowing the system to go lower in the bass.
The ES14 should use a porous bung in the port. Fully open and fully closed is wrong.
Open cell speaker cover foam is good for the job. Roll it lightly into shape and pop into the port.
Indeed the ES14s sound very even and open across the whole frequency range.
They do and that is a tribute to the skill of Robin Marshall.
Tweeter is the week area IMO, though it is still good. Small domes don't like going too low on such a shallow crossover slope. Robin stated that his aim was to use a good ribbon on the ES14 at some point.
Well a true ribbon might not work ideally but a nice isoplanar might
