Correct! And the one beneficial mutation is more likely to persist and be passed down, possibly causing morphological change.
The last point of your post was sheer lunacy. I'm sorry, but I'm tired, and at such times my opinions become more unvarnished. I did predict you'd attempt to implicate water, however.
There is a difference between applied science and pure science. Applied science is about making money, such as developing new drugs and high price treatments. Applied science employs methods that are fast and cost effective so one can compete in a fast moving market place. Statistics is useful for this purpose because it does not have the requirement of having to pain scathingly predict first, using logic, before you can try anything. The logically approach is a pure science method, that requires all the ducks have to be in a row, so a magic eight ball is not used to cover blemishes. If the goal is to make money, applied tools are used.
What tends to happen is because good drugs and good treatments constantly come to market (employ the majority of life scientists) one assumes this is because the methods used reflects pure science. Pure science is too slow if the goal is to make money; tools are used to speed up the process.
The analogy of applied science is a good chef does not have to have a degree in food chemistry to make a soufflé rise. He can do this with trial and error and empirical observations, while completely ignoring details that occur at the chemical level. This efficiency is what applied science is all about. A pure science approach would start with the food chemistry and predict what is needed to make a soufflé and get it right the first try.
To make a new food texture, the empirical chef will have hunch for a starting place based on experience and then try a bunch of things. This will be faster that having to define the chemical logic before you can run the proving experiment. When you develop new drugs you don't need to include water, because the applied approach of the chefs lumps this empirically. The assumption is water does not matter and can be substituted like vanilla or chocolate. The assumptions of this simplifying approach, although useful to expedite products to market, is applied to evolution; water can be substituted and therefore ignored. This topic does not allow me to deal with physical chemists who understand the significance of water and organics.
In terms of chef talk, the idea of one beneficial mutation being passed down and selected will work if there is some unspoken rule of nature that says that is can only be 1 mutation per person, at a time. If there are two or more mutations per person, the odds are the second and third will be bad, not good. If you have a bad mutation and a good mutation in the same person, this is not natural selection but natural attrition. Selection is the best of the best not what is left over. This process of selection is good for drugs.
The analogy is we have 1000 men standing around with a gun that can shot 1000 bullets. We take one bullet out and require each one put the gun to their head and pull the trigger. The last one standing is selected and gets all the spoils of the other 999. This is not a very good system. It may work to make money but it defies pure and natural logic.