To begin, there is currently no gauge that measures neck tension. A 'strain gauge' does not. Our seating force gauges do not.
The term 'tension' applies to grip, and not friction, and not interference fit. It is bullet grip that affects pressure.
Grip is purely a matter of spring back applied to the seated bearing area of our bullets.
Anyone can test this on the bench,, I just did with some new-brown box 6br I'm getting ready to turn, to provide some numbers.
It's 13thou thick at the necks X2 + .243 is .269"
I 265 bushing sized and get an unloaded OD of .264" (-5thou)
I seat a Berger and confirm loaded OD of .269"
I then pull that bullet and unloaded OD is ~.2688" (I can re-seat with force of hand)
This was 5thou of sizing, and you need to understand that tension never exceeded spring back, which was near zero in this case, because it's with a new/annealed neck. There was plenty of seating force with that interference fit, in squeaky clean neck, but seating force is meaningless here.
To raise tension I could cycle size to harden the necks. It takes a good number of cycles at this sizing amount to raise spring back even to ~1thou. 50cycles might get me ~2thou of spring back,, maybe. So where it's suggested that frequent annealing is needed, I have to assume gross oversizing(like some of it suggested around here).
I should repeat; Grip/Tension is purely a matter of spring back applied to the seated bearing area of our bullets. The greater the spring back(1-3thou max, ~1thou normal), and the area applied, the greater the tension. This can be manipulated, controlled, and measured across a chrono.
I do it by managing my brass hardness, and adjusting LENGTH of neck sizing(not amount). My length of sizing usually remains onto seated bearing.
It is possible to increase 'effective' tension beyond actual spring back, by sizing more neck length than seated bearing in it. It's effective in that the bullet is not released until all the stored neck energy is overcome. This is the only way downsizing beyond spring back can still affect tension. But keep in mind this is effective tension, and not actual. Much like seating bearing into donut can increase effective tension, regardless of sizing amount.
I don't believe it's right to imply that reloaders should blindly go with sizing in excess of spring back. I'm confident that in most cases, by far, excess sizing provides no advantage, but detriment.