My build log

[Jonathan]

Alex, as Martin says I'm sure the 16mm rail will happily do the loads but the 20mm has an extra row of balls (5 instead of 4 I'm led to believe)

I know SBR25 has one more row of balls compared to SBR20. Not sure about 16mm. Either way you would be much better off getting 20mm (or better) 25mm rails. SBR25 for the 1500mm rails and SBR20 for the rest would be good. It seems you're using 4 rails on the X-axis, which has clear advantages but will make it difficult to align all the rails such that they are parallel.

You will definately need 2 ball screws as the gantry will rack badly at 1000mm wide and you could end up damaging your machine as well as your work piece.

Yep...

I'm sure Jonathan will chip in at some point soon and will probably recomend 1610 ball screws instead of the 1605 and using belts and pulleys to improve the resolution whilst still keeping the speed.

Yep... on X and Y you should use RM1610 since this halves the rpm the ballscrew must do to get a given feedrate, so if the feedrate is limited by the critical speed of the screw (which it certainly will be over 1500mm) then you get the higher pitch screw to compensate. Read any of the build logs on this forum and you'll probably find an explanation of this.

200mm is quite small for the Z-axis, but in a way that's good since if you can cope with having so little travel (50-80mm ish) it makes it easier to design a strong machine.

Once we have the set up the way we want it, and its time for the electronics, we will be ordering them all built up and ready to plug and play

Waiting until the end is a good idea since it means you don't risk wasting money if your objectives change, however I'll be surprised if you can find a plug and play system that's worth buying. Most of them seem to be very mismatched, or highly overpriced.