The band is removed, but not the wedge, which is tightened immediately. I dont usually see bleeding here.
Petroleum jelly is only used on the adjacent composite contact area if it has just been placed. If used judiciously, you should be able to avoid contaminating the composite. It does not have to be used if the composite is 24 hours old, or if it is a natural tooth, porcelain inlay/crown, gold inlay/crown or an amalgam.
Having the patient bite into the uncured resin shouldnt contaminate the field with saliva if you are working in a dry field, which is mandatory in curing procedures. I havent found sticking to the opposing tooth to be a problem. If you are finding this occurring, before biting into the uncured resin, trim the excess and place the occlusal anatomy, light-cure and then start to adjust the occlusion.
There should be no or very little overhang if your band has been wedged at the gingival margin and the band is touching the adjacent tooth in the contact area. You have left only 1 millimeter to be filled after the band has been removed. The band has given you the contour. If you cannot achieve this, you should use an alternate technique.
This technique is completely amenable to the rubber dam technique, which I highly recommend. The final curing would be done with the rubber dam in place, and the occlusion would be adjusted after the rubber dam is removed.
I developed this technique after trying many of the commercial matrix systems on the market and not finding any that were completely satisfactory. Developing tight contacts in Class II resin-bonded composites is not a simple procedure. Employing my technique will give predictable tight contacts, but it does require some practice to master this method.