What is the average spread of F.P.S. in a good load from shot to shot? Should I expect a spread of 20,30 or say 100+ FPS? What is normal in a top load from shot ot shot?

Single digit Standard Deviation is optimum. SD's in the teens is good. Marginal in the 20's. Yet with all this I've had loads with horrible extreme spread (near 100fps)that shot better at 100 yards than the single digit loads. The single digit loads, if they can be tuned to the gun, will give consistently better accuracy and be more reliable in the long haul.

I Shooter: I save all my chronograph readings in a looseleaf binder. In addition many years ago I used to sell my chronograph services whenever I was at the range! I have seen many 5 shot runs of bullet velocities go 125 FPS in variance! And these large variance guns sometimes shot good groups at 100 yards! My friend Louie ran five shots over my Oehler chronograph one day and his extreme spread was 130 FPS but the group was just .750"! Who knows what the group would have been at 400 yards but at 100 yards that was good shootin! I do not worry one bit when my extreme spreads on higher velocity Varminters runs 50 FPS! Chronographs do tell you one very important thing (most important among all the other things it tells) - your exact average velocity and of course you can figure your long range drop tables precisely from there. One should of course strive to use a powder and primer combo that gives consistant velocity but again I have some Rifle/load combinations that give 75 to 100 fps extreme spreads and they group well and perform well in the Varmint fields. If I found a load that gave good groups and was a bullet that I wanted to use in that Rifle I sure would not mess with it anymore even if it gave 75 FPS extreme spreads! Chronographs cause questions!
Hold into the wind
VarmintGuy

I usually find a standard deviation of about 35 fps or so in commercial ammo. Without a lot of effort, you can get low 20's. With a bit of effort you can get into the teens.

There is a lot of debate over the actual "point of diminishing returns". It's just an opinion, but I think that if you're in the low 20's, you're probably going to be about as accurate as you're going to get.

I can tell you for sure that 5 shots do not establish a precise standard deviation. It takes several dozen shot to get much precision. 5 shot samples from the same lot will produce 2:1 differences in standard deviation, just because of sampling error.