1) Confluency is something you learn with experience. The best way is to get someone to show you what they consider 100%, 70%, 50% confluent for the cells they are working on. There is a bit of variation with different cell types, and some types such as neurons are very tricky to work with, but practice is the only way to go.
2) There are a ton of posts on here about this very topic: have a look at the attachment in my reply to this topic: http://www.protocol-...ing/#entry55266
3) Just like making dilutions for a solution - C1V1=C2V2. Just make sure that you are resuspending your cells fully before making the dilution, and while plating - they settle out fairly quickly.
4) This depends on the transfection reagent you are using. Check the manual for the reagent you have for specifics. Generally the process involves making a dilution of DNA in medium (often serum free), then adding transfection reagent (volume based on amount of DNA), mixing, incubating for a few minutes, then adding to cells. Some protocols/reagents call for making a medium/transfection reagent mix, then adding DNA.