The runout is specified in the bearing class (ABEC spec Numbering) along with the clearances. With a pair of bearings preloading one against the other takes up the clearancees leaving only the run out. In the pair used it is possible for the run out to be in oposite directions which will cause some initial wear and possibly shortened life. If the shaft and mounting are stiff enough possibly even a detectable rotational variation in turning torque might be felt as the preload is applied - however this will require some good close tolerance force fit type machining of boththe shaft and the bearing housings. However with modern bearings of know manufacture this is going to be rare.

Recently manufactured ball races using cages (i.e. no filling slots) can support a thrust load that is more than enough for most applications at the speeds met with on a lead screw. For higher speed angular contact bearings are required but are over kill and if the thrust is a consideration then separate thrust races should be used as a pair to accomodate end thrust and a single ball bearing used to provide location as a slip fit on the shaft or if the thrust is bigger still then taper roler races are the way to go on a big machine.

If you want a real cheap and quick bearing look at the 5000 series dual row ball bearings - these are pre loaded and are available in a range of stock sizes and low prices from all the reputable makers. The cost is less than half the cost of a pair of normal ball bearings so represent good value as well as being available with seals fitted to keep then clean.

The linear location tolerance along the screw is what differentiates C7 from C3 with C5 grade also available. The C3 are better and would normally be used on higher class machines the accuracy of the end support bearings is by comparison not going to make any detectable contribution to the accuracy measured along the lead screw if using a pre loaded pair. The AC bearing might give a greater thrust capability but the limiting factor will be the number of balls in active contact with the screw. Look to the manufacturers tables for the bearings you have and you will see what I am driving at. Make sure the tables are for the actual make of bearing you are using as there are differences! If the bearings are unbranded then I would halve the table values but you should find there is plenty of thrust capacity and expected life. If you do the calculations then you will see that it is the constant reversing that will reduce the life more than the number of revolutions.

Regards

Pat