The Nomad is quite a stable, rigid machine - 0.001" repeatability. Small doesn’t mean lack of quality.
I can’t speak to the SO3 but I would suspect it doesn’t do as well as that Nomad for repeatability.
I find the chip load tables to be less useful for these less rigid machines.
I find the chip load tables amazingly useful. I use them to double check what G-Wizard is telling me.
Can’t agree more about needing to experiment, regardless of tables. I teach chevrons, rather than spirals. YMMV.
Metals and plastics, as long as one knows the formulation/alloy, perform very close to their tabled data. I rarely find the need to experiment; the tables and G-Wizard are “spot on”.
Wood on the other hand, absolutely requires experimentation. For example, Oak. Quercus Alba (American)? Quercus Robur (European)? Which area? Oak can vary in density and hardness by quite a bit. Pine? Is it Yellow Pine? Yellow Pine is a hardwood, not a software or medium wood.
End mill diameter is a major factor in machining. It affects speed and finish. Flutes? There are many to choose from, each with interesting results. More flutes doesn’t always mean faster… and finish may not be as good. Noise? Yup! A proper enclosure and this isn’t as much as an issue.
Good feeds and speeds is not just about the simple calculation. There has to be sufficient torque to keep it going. One also have to add the deflection adjustment for the “stick out”. When all is done properly, the results are as good as it can get.
Using others feeds and speeds is likely to lead to disasters. Unless from a trusted source, one could have poor numbers, really mistaken numbers, or numbers that do not work for you (stock variation).