A properly tightened head bolt acts like a spring (applying clamping force). You tighten close to or slightly into the point of yield (permanent deformation, i.e. elongation). The 90° beyond is guaranteeing yielding of the bolt. (It is calculated not to yield it too much, as going beyond initial yield just results in necking down and weakening that spring effect and concentrates it thru a smaller cross-section.)
(“Normal” bolt torque specs. are to some specified percentage of yield, for example to 75% of yield.)
Torque is not appreciably increasing thru that additional 90° (because the bolt is yielding), which is why the last bit of tightening is not by torque, but by distance (degrees). It is ensuring that you are into the maximum spring action that is needed for adequate clamping, but not into a neck-down on the way to “ultimate yield”, AKA, the breaking point.
I don’t claim to have perfect understanding of it, but that’s the best I can explain it. As an afterthought, I looked it up: https://en.m.wikipedia.org/wiki/Torq...yield_fastener
Regarding the lubing, any valid torque spec. should be with stated or assumed lubed or non-lubed thread. I believe in industry, unless otherwise stated, it is assumed that the threads are lightly lubed. The assumption may vary within specific industries. Torqueing wheel lug nuts is an exception - they are torqued to spec. dry unless otherwise stated.
Lubed or not lubed makes a difference. The real goal on any torque spec. is the resulting clamping force. That determines where you are (again, as percentage of yield) for the bolt’s material (including tempering) and cross-sectional area. If you lube a bolt for which the stated torque spec. is for non-lubed condition, you decrease the margin to the yield condition, or perhaps could yield and weaken it. Doing the opposite will result in lower clamping force, which could be a problem in critical applications (cylinder head).
In the case of torque-to-yield head bolts, for them to be specifying that it be lubed, apparently it was determined that the end effect (clamping force and being reliably close to the yield point) is more accurate and consistent with it lubed at a lower torque than lubed at some higher torque value that would in theory give the same clamping force.
BTW, with stainless steel fasteners, it is pretty much imperative that you use anti-seize on the threads. Stainless steel threads will gall (cold weld) without anti-seize - IOW, before you reach the specified torque value, the bolt and nut threads will be irreversibly fused together, and you’ll have to break or cut the bolt off and start over with new hardware. The anti-seize will have its own coefficient of friction different than oil, so that has to be taken into account in critical applications. Often where stainless steel fasteners are specified, the type of anti-seize will also be called out, and the specified torque value will have that taken into account.