Generally antibiotic for gram positive and gram negative bacteria are different. Good antibiotic for gram positive bacteria are: Vancomycin, teicoplanin, linezolid, quinupristin etc. For gram negative bacteria good antibiotics are: Cephalosporins, Fluoroquinolones, Aminoglycosides etc. However, you may try for broad spectrum penicillins with or without beta-lactamase inhibitors for both the cases. Of-course it is generally good for gram negative bacteria.

Yes.

I think you shoud use traditional plant antimicrobial peptides as posotove control in your research, like defensins, thionins or lipid-transfewr proteins. It would be more representable. In my group we study peptidomics of Taraxacum officinal Wigg. If you're interested, I'll be able to send you our papers.

Ok, very good and interest work

If I understand your query well, it is better to use different positive control for Gram (+Ve) and Gram ( -Ve) bacteria. If I were you, I would have add another additional control of broad spectrum antibiotics or plant peptide as suggested by Eugene.

Your choice of antibiotics to be used as control should depend on the test isolates selected. You need to be sure that the antibiotics are not those at disadvantage to your test isolates. If your test isolates are carbapenemases producers, the choice of any beta lactam antibiotic would be wrong. I would suggest a broad spectrum antibiotics belonging to the third generation fluoroquinolone, provided your isolates do not have efflux-pump mechanism

You can use one broad spectrum antibiotics for both the organisms keeping in mind to maintain a reference paper of the positive affection of that antibiotics on your selected organisms. however your main objective is to test the bioactivity of plant extract.