I figured it out. I was not clamping the knife down with the top screw all the way down first. On the jig it says, “1” on the big knob, the “2” on the front screw, and “3” on the other side of the black knob. So, I was going in those steps to clamp it, which is very difficult. I always thought that was odd…Anyways, someone mentioned holding the clamp tight on the knife, tightening the front screw down as far as it would go and then tightening the black knob. Did that and it eliminated the angle difference. Phew!!
...snip...
Yes, these jigs are essentially parallel clamps and they work best when the workpiece is clamped as parallel as possible. Hard to tell when many knives have such narrow spines that are unground, parallel steel.
Yesterday, motivated by your issue, I started doing some testing, with a surface plate, gauge blocks, calipers and a height gauge, but I used a flat piece of steel for making measurements rather than an actual blade. After a lot of work, I simply demonstrated to myself that despite my perceptions of how well a blade might be clamped by looking at the gaps at the front of the jaws, the blade is held the most parallel, in the same plane as the jig centerline, when the top jaw, behind the front screw, is the most parallel with the lower jaw. (If prompted, I will inflict upon everyone all the gory details, tables, graphs and pictures, so be careful what you ask for

My take away is that one should open the screws when inserting a blade, tighten the front screw until it gets a little snug and then tighten the rear screw. And then check how parallel the inside jaw surfaces - from the are, by eye. Adjust BOTH screws, as needed, until the jaws are as parallel as you can get them. Don't worry about gaps at the front of the jaws where the blade is held, just being sure it is held securely enough that it won't shift during use. (Set screw used in place of the large knob for these tests only)

I think what texaspro posted about holding the knife in the jig jaws and then tightening the screws in order is a good general approach. I think it doesn't go far enough because it doesn't include checking for and adjusting for parallelism of the jaws.
One noteworthy observation is that it is possible to bend the top jaw with the rear screw, which I think is unnecessary. Only use as much pressure as needed to securely hold the blade.
Rick