Mods can increase drag if they aren't right.
Here are my $0.02 on your mods:
- Rear wheel skirts look good except maybe the gaps. I have some skirts with even bigger gaps, my feeling is it's still a net positive, but not ideal.
- I'm not sure anyone knows, certainly I don't know, but I would bet that an air dam bending inward can increase drag. If the air is not stopped by the dam, but continues under the car, then the result is a larger number of messy interactions. A camel with one hump is better than a camel with two humps. A stiffer dam that is angled forwards if possible, is probably better aerodynamically. Not only does a dam need to stop the air, the air needs to end up going over the top or sides of the car - that's the whole point. I'm worried about your air dam, especially if it's not stiff enough.
- Others mentioned bending your Kammback down a bit, I would also say it would work better with some sides, but I'm still guessing.
- I would suggest (partial) grill blocks next.
Unfortunately, I think that you either have to copy tried-and-true designs, or else treat it as a research project:
- Mockup in cardboard and tuft test, if possible. Differences in audible noise levels also provide clues. A-B-A coast-down or fuel consumption instrumentation tests are better but take more time. CFD would be ideal, but it's expensive and most of us don't have access.
- Fabricate.
- Setup a definitive test to verify efficacy, if A-B-A wasn't done in step 1. It's fine to group mods together, more especially if you have confidence you know what each one does. If the magnitude of the improvement is large enough then tank logs or a known road trip and weather monitoring may work. Statistics can help.
- If the mod didn't work as expected, go back to step 1.
(EDIT: Actually, even if you pick a tried-and-true design, but if your car has a different starting shape that might affect the mod, it may still be a research project.)