How much ease is allowed beyond the basic chest measure?
The width from CF to CB is half the chest circumference (taken on the skin) plus 2". So that's 4" ease over the full chest.
the neckline is off.
Too low in front and too wide in the back. figure out why- is the draft bad or is there a mistake in reading the draft?
Back neck width is 1/12th chest + 1/4". No seam was added to the neck hole, so what you are seeing is the location of a finished seam. It's almost touching the neck, so should it really be smaller?
As for overall distribution of circumference, I think there is just a bit too much fabric in the back and not enough in the front for you.
These are based on proportions of the chest. On the back, scye to CB is 1/6 chest + 1-1/2" On the front, scye to CF is 1/6 chest + 1". These were within about 1/8" of my wife's measures on my body, but she may not have measured accurately.
How to determine how much to add or remove? If both the back and back neck are too wide, should I try taking in the back seam?
From the back photos , you can see (especially on your left) how the fabric is slightly taut over the upper blade. Need length to go over that part of the body. So unpick the top of the sleeve. and unpick 3/4 (or most of the shoulder seam length). Leave the shoulder seam allowances pressed as they are. You should baste in a piece of muslin under the front shoulder to act as an inlay.
Get help for this next part.
Put the shirt back on and get your helper to pin the shoulder seam closed where it wants to naturally lay.
You will likely need to pin a dart in the back shoulder first, then pin the back shoulder onto the inlay you basted in place.
Do the left and then the right, which will be different as your right side is dropped. Don't try to fix the drop yet.
Is this essentially the same as the alteration posaune suggested?
If so, I think I would be more comfortable guesstimating the extra length needed, and adding a separate back piece, which will be needed anyway. My wife is not a skilled fitter, and I would rather have a controlled, repeatable change than a series of random experiments.