Imagine a hypothetical individual creature, PERFECTLY adapted to its environment as a result of its genes. This perfectly adapted creature has offspring, the genetic replication process is imperfect and errors occur.
By definition 100% of these mutations will be neutral or disadvantageous.
When you are at the summit of a mountain any movement either way keeps you at the same height or lowers you.
Now these children have children who again have mutations. It may be true that there are more ways to go down the mountain than up it (the lines converge as you go up, that is: the amount of area immediately up is less than the amount of area immediately down - another way of saying there are more ways to break a well designed machine than improve it), but by chance some offspring will be closer to their grandparent's 'perfect state'. If it transpires that offspring higher up the mountain of perfection have more offspring, then we would expect that the population would climb and hover around the summit. Those that are born too far down, don't have children that have children nearly as much as those above, and struggling for the same pool of resources, eventually their bloodline goes extinct.
Imagine the creature was at the bottom of said mountain. It's able to survive long enough to have children, but only because the competition for food is weak. However, at the foot of the mountain it might be the case that half of all offspring have mutations that are advantageous - there's much more mountain, there are more paths available that lead up.
Evolution is something like this with complications. The mountains have thousands of dimensions, and are constantly shifting in height based on what other mountains are doing.
And so, it is difficult to say whether a certain fossil has a beneficial or disadvantageous mutation, not just because generally only hard features such as bone are preserved - but also because we don't know the shape of the mountains. We don't know the exact challenges faced by the creature, and we can't compare it with the population at large.
How many mutations are beneficial is determined by how 'fit' the population is to its environment. The more specially adapted it is, the easier it is for small perturbations to disrupt the adaptation. An example: Humans are built under the assumption that we'll have four limbs. It is possible to live with fewer, but even with communal assistance, the survival/birth rate is likely inhibited in the very long term. Trying to add more limbs is even worse. Our physiology is 'fine tuned' by evolution {or God if you insist} so any mutation that results in extra limbs is likely to cause problems elsewhere even if they are improbably useful limbs. Blood flow management, the nervous system, balance and so on are all likely to be impaired by the new structures.
™ Version 4.2
