Regarding the variation or not of EOT with longitude, I can see that if EOT is taken as the original meaning of difference between (local) mean solar time and apparent time (i.e. local solar time), then it MIGHT vary - for example if mean solar time was calculated with reference to local noon, as the time of noon would vary with longitude - although I am not clear in my mind whether it actually WOULD vary. It would depend, I think, on the period over which the "mean" of the "mean sun" was taken.
However, once one has fixed on a particular prime meridian (Greenwich) for the location of the mean sun's observation, and reduced other longitudes with reference to that meridian, EOT can automatically become a worldwide quantity, as shown by the definition of EOT given earlier as (local solar time) - (UT adjusted for longitude). This remains true regardless of how mean solar time is calculated, given that UT is based on "atomic" time with a small correction tying it to astronomical time. Please can someone point me to an explanation of exactly how the mean sun is derived? ... but anyway these effects are too small to see with a sundial, aren't they! ? Regards Andrew James