Civilization – Worker automation, can’t fine tune as much as in Civ4

First of all, you can just sell them if you don't need them. Not only do you immediately get some gold, you also decrease the number of units you own, and units cost gold each turn.

Secondly, connecting a city to the capital by railroads greatly increases a city's production, so if you can afford it it's recommended.

Finally, you can replace improvements you don't need / want with other improvements. For instance, for cities that are big enough it's probably better to replace farms with mines or trading posts.

You can set your workers to "automate", and they will automatically build roads, replace them with railroads if available, and will replace improvements with other improvements the AI thinks you need more.

You can click on an automated worker to see what his current orders are (it's highlighted), although if that worker is being moved across the continent to build something over there, you'll see only "Move", not "Moving to [x,y] to build [z]", which is unfortunate.

That is, to the best of my knowledge, the closest you can come to finding out what an automated worker is "planned" to do, and it's not really all that great.

FWIW, I, too, have noticed that automated workers often bunch up in one area, and then seemingly stop doing anything because there's nothing else to do in that area, ignoring other areas (especially those I've recently expanded into) that could still benefit greatly from improvements.

I use a couple of different techniques for addressing these situations:

1. When I build a new city/conquer a new city, one of the first things I do is build and then automate a worker. I end up with my capital city supporting 2-3 workers (start of the game 1 is just not enough, but neither are there enough cities yet), and then every other city supporting 1 worker each (although sometimes I simply change the home base of one of the capital's workers instead of building another, usually when I already have a sprawling empire built up). Anyway, the takeaway here is that building and then automating a worker usually (although not always) results in said worker improving the local area first, and this also seems to oftentimes trigger the automation AI to realize that work needs to be done there and send idle workers over to help out.
2. The other trick I use is to de-automate 1-2 workers, walk them over to where I want them to work, and then re-automate them again. Same principle as the previous -- they tend to work the local area first, and often trigger help coming from other idle workers -- although it seems (from my entirely subjective point of view) to work less often than the other method (they often just walk right back to where I just took them from to stand around idly again). Walking them over, manually building something, and then re-automating them seems to work quite well, though.