All satellites are in geostationary orbit at a fixed "height" above the Earth, in a ring around the equator, called the Clarke Belt.
Seen from the surface this means they lie on an arc that touches the horizon at about 75deg West and 75 deg East (but this depends on your location). The elevation - the height above the horizon of satellites on that arc - depends on your location as well. The further away from the equator you live, the lower in the sky they all appear. If you lived on the equator, the highest point of the arc would be directly overhead. From the UK (you don't say where you are, but let's assume it;s the UK) - the highest point of the arc is about 30degrees or so above the horizon.
And since the sats are on an arc, then yes, each sat is at a different elevation.
Here's a link to a very useful program which can calculate this for you if you tell it your longitude and lattitude...
you can manually set the skew (angle of orientation of a magnetic or mechanical polarisor) on many high-end receivers. This can be useful especially in getting the sweetest spot of a circularly polarized signal, but for linear signals it's hardly useful. That's because as the dish tracks the arc, and the LNB (incl. polarisor) consequently tilts in unison, things contrive to keep it perfectly aligned with the skew of the sats further along the arc. If you were to aim a fixed dish directly at a low-arc sat, then you would need to rotate the LNB (set the skew) to match the settings you refer to.
The link mentioned (http://www.smw.se/smwlink.htm) earlier in this thread says that Astra (I presume 19.2E?) uses a non-standard polarisation angle and indeed the program gives values that do not follow the expected trend.
What does this mean to a motorised dish, presumably one should set the skew using a non Astra signal?
Does an incorrect skew angle make a lot of difference compared to other factors, like LNB and dish size?
yes, Astra1 has a skew that would normally be found on a sat at 8 deg.east. Don't ask me why.
The 7 degree angular difference between what it "should be" (-2.6deg) and what it is (4.3deg), is not very significant for Astra1. For one thing, there are no circularly polarised signals there. For another, the signals are so strong that even with a grossly misaligned polarisor, there's plenty coming through.
You can experiment yourself here.
Try rotating your LNB in its mounting. You'll find that it quite happily copes with a misalignment of almost 45 degrees...