All brass is tapered from the mouth to the base. Measure any brass, regardless of it's origin and you will find the neck is thicker at the base than at the mouth. Further, you will find the brass in the shoulder is no thicker (in fact, it's usually thinner) behind the neck than the neck is at the base. This isn't a theory, this is how the brass measures. Everybody's brass. When you neck up a piece of brass, the brass in the neck thins out, that in the shoulder is unchanged. I don't have to guess or theorize about this because I have measured brass ising a .0001" tubing mike to find out. It was true of the 220 Russian Sako brass I measured in the '70's, it's true of Lapua 243 and 308 brass measured today. The brass shown in the picture above is typical of what you will see with any brass. The neck is thicker at the base than at the mouth.
Now, if you have formed brass from a longer or larger case and reamed or turned the neck, the neck will now be thinner than the slope of the shoulder. After a few cycles of firing and re-sizing, the thicker brass will move into the base of the neck and you will have a doughnut.
By the way, I have yet to be able to show that brass flows significantly, regardless of shoulder angle, from being fired. Brass flow is primarily the result of resizing. For those who doubt this, try firing and reloading a cartridge with a sloping shoulder (I used a 30/06 and a 303) sizing just enough of the neck to hold a bullet. With the 30/06, I got .006" of case stretch after twelve firings. I was sizing 2/3 of the neck and I suspect most of the stretching was in the neck only. With the 303, I sized only about 1/16" of the neck and a dozen loadings gave me .0025". This was in a P14 actioned rifle. A lee Enfield will stretch cases as a function of the action. With either cartridge, if I full length re-sized just once, I got up to .006 of "flow".
Back to the doughnut. If a person wants to turn his necks and has enough clearance to allow it, he is much better off to fire the cases a couple times before turning. This does a better job of ironing out the wrinkle which results from necking up 243 to 260, for instance, than an expander ball and sizing die does. This is why I recommend using a blank load to expand necks. I have seen instances where guys have sized up brass and not re-sized the entire neck. Then they turned the necks and ended up with necks which were thin at the base because the turning tool pilot was supported by the sized portion of the neck and turned the unsized portion to the same OD as the sized portion. This resulted in what might be called a negative doughnut. Probably harmless.
Once again, the brass in the shoulder area is no thicker than the brass in the neck unless the neck has been turned or reamed. If you blow the case out straight, you will have a case wall which is tapered from base to mouth. When it is sized to form the neck, the neck portion thickens a bit so that it actually becomes thicker than the shoulder.
By the way, partly because the brass in a cartridge neck is tapered, standard chamber specs also call for a tapered neck (it also reams cleaner). Sizing dies generally have a straight neck. Custom chambers may have a straight neck (most benchrest chambers always have had) because it is much easier to work with if trying to fit necks and/or maintain consistent neck tension. The problem which arises is that, once the brass is machined to consistent thickness, it is now thinner than the shoulder area and a doughnut will inevitably, form.
The old Lee Target Model Loader used a neck sizer in which the sizer was tapered and the brass was then bored with an eccentric reamer. As a result, the outside of the neck was tapered yet the inside was parallel. Not a bad system as long as that sizer was used because the dimension of the inside of the neck remained constant. If a doughnut did form, it was a simple matter to bore it out and retain the proper dimension for the rest of the neck.
The bottom line is this. It matters not whether the brass is expanded to a larger caliber or sized to a smaller one; the brass is going to resist forming. As a consequence, I think it is a good idea to shoot the brass a couple of times before doing any neck turning. Regards, Bill