It sounds as if you are less worried about the mechanics of project, and more concerned with showing whether progress is on course- a god thing, in many ways.

As far as showing progress goes, you need to define the parameters of what you want to compare- you mentioned overall % complete and task variance. My company has been going through a period of reviewing progress mapping, and it has thrown up a few things.
As far as overall progress goes, this often does not show a true picture, especially if you are using TRAs (which to my mind, a good programme should do). Reducing the TRA proportionately to risks allocated to the TRA helps, but can lead to false reporting during the first half, with additional delays in the second half causing big increases in completion.

Tasks started on time is another indicator, but again if one task starts late, all following tasks (if the programme works) will also be late, so this tends to be misleading. Task finish dates can be a bit more useful, but again it doesnt take much to throw things out and cause misleading results.
This is reflected in your task duration variance- early delays can throw the programme out, but if succeeding tasks take the allotted time, then reporting will be good even though the project is on a late finish. If you want to see what areas are performing, then this can have value, but apart from that it needs to be taken for what it is.

As an overall indicator, Planned Project Completion is probably the best indicator. This is a one item, easy to understand indicator that allows you to see immediately if you are on or off track.
There is a system I have used for a client where the project % complete of critical path items, as an overall number (your summary bar) is compared to the %age of TRA used, where TRA is indicated as a single activity at the end of the programme. This allows for a direct comparison of overall progress, but also allows for allowed slippage. As an example, if you have a project of 100 days and a TRA of 10 days, then if the project is 50% complete and you have used 5 days of the TRA (50%) it would be deemed acceptable. this allows for overall delays as opposed to individual items, and can give a more complete picture.

I may have gone completely away from the subject, but as far as reporting goes, the main thing is to decide upon the importance of each possible indicator, and whether it is relevant to what you need to show, If you can get that, the actual reporting tends to be both straightforward and useful.