My experience is that the weights of the components in Rocsim are usually correct. The biggest problem I run into is things like glue, paint, and the dozens of little parts that are just inconvenient to include in the design screen. If you are building a sport rocket, then you just build the rocket and add nose weight until the actual CG is one caliber ahead of the CP as predicted by Rocsim. If you are looking for max performance, then you don't want to be adding nose weight and will want to develop your own database (based on experience) as to how much weight the paint, glue, etc., add and where to locate those additional masses.
The one caliber guideline is a generality and may not be appropriate for all rockets. At zero angle of attack, a circular cylinder gives no aerodynamic contribution to stability (or instability) and hence does not enter into the calculated CP. As the angle of attack increases, the cylinder (body tube) does begin to contribute and most often it is a destabilizing influence. The one caliber margin accounts for this and any inaccuracies in the calculation procedure. If your rocket is long and skinny, the impact of the body tube will be greater and you will need more than one caliber margin. If it is short and fat, it is entirely possible that the aerodynamic center of the body tube is behind the CG and is thus a stabilizing influence. In that case, you can get by with less than one caliber margin, but don't push it too far. You will find that the LOC Mini-Magg is such a rocket. People fly them successfully with less than one caliber of stability. I have actually seen one that had a large engine and was unstable at zero angle of attack and became stable as it yawed. It zig-zagged all over the sky. This is not a desirable situation.