I think the solution would be to have each item useful in certain situations, but there would be different qualities of each item of the same type. If you had 3 different types of axes, each would be better at specific tasks, and different axes of the same type would do more or less damage or cut trees faster or slower dependent on their individual quality. Quality would be dependent on the materials used to craft it and the skill of the crafter. This would create a demand for all different types of tools and resources to make those tools. Then not everyone will be looking for the uber tool, because it will only be uber in one instance and not another. People will find the best tool they need for the task at hand. I could use a sledge hammer or a claw hammer to drive a nail, I will choose the claw hammer, but when breaking a rock, I will use the sledge. Both useful in both jobs, but clearly one or the other is a better choice for different jobs. The research skill sounds good, but it should give increasingly more verbose and descriptive details about an item you are using or crafting the more you handle or craft it, just as an expert will know more about his tools than a neophyte. With details about an item, you could then use that information to choose the right tool for different tasks. That will give us the same type of visceral cues we use in real life to make decisions instead of just a number. If people have a fear that tools will be benchmarked to find which is best in a given situation, there should be enough disparity between them to make the choice blazingly obvious. Similar tools should be so close to the same to make it just a visual choice.