The path that Microsoft is going says yes. They are moving all there samples and defaults to using uni-code. In reality uni-code may not be needed in your environment. If you are storing only Latin characters (the US alphabet characters) then you probably don’t need to store uni-code characters. However if you are planning on moving your operations to support customers who do not speak English then you will want to setup your database and application to use the uni-code data types (nvarchar, nchar, ntext).
It doesn’t take much extra to use uni-code data types, however do keep in mind that it does require more disk space to use uni-code data types. For each character saved two bytes of disk space are used instead of one byte. While this may not seam like much space if you will have a large table with a large ntext field this can use a lot of space quickly.