These may be too late to count now, but a few suggestions in case:
- Low gear, elevated engine speed, light load, low EGT are the best recipe for keeping cylinder head temperatures under control. I drove my 745 TD over the Grapevine last July, did it around midnight, took it easy and had no problems. I found later in that same trip climbing up other passes that I could make the needle start to climb by selecting a higher gear and putting my foot down further to maintain the same speed. 3rd gear at around 3200-3500 rpm with no more than 3-4 psi of boost on the gauge kept it in check even when ambient temps were over 105F. Pushing any more than that in those circumstances led to issues.
- Cooler thermostat is a huge factor and adds greatly to thermal stability and engine's ability to exhaust heat. 80C Wahler t-stat, part # 035 121 113, is the best one and makes a major difference in engine's ability to transfer heat to the rad.
- I think your premise that the engine will not suffer as long as the coolant does not vaporize is flawed. I think devoting a lot of energy to knowing exactly how hot it the motor is and how hot it is able to get so that you can run as close to that threshold as possible is also not a great approach. For one, we don't know 100% that the head can't be damaged as long as the coolant is liquid. We do know that the head doesn't like to get much hotter than normal operating temp and that headgasket failure threshold conditions are unusually easy to achieve with this engine, relative to others. For another, your ability to really know cyl head temps is limited since you're only measuring at one (stock) or at most a few points -- temp could be higher elsewhere, up to and beyond the boiling point, even while according to your measurements you're maintaining a >5C distance from your calculated limit value. Your decisions could be affected by incomplete or inaccurate information, and something unexpected might happen at a point you thought would be safe. Also in terms of unexpected happenings, to keep the coolant liquid requires running system pressure as high as possible, which involves increased risk of a component rupturing. Combine the intention to run the motor at temperatures well in excess of target normal operating temp -- leaving a limited cushion for reaction if something happens that causes temps to suddenly rise further -- with the necessity of high pressure and its affiliated risks, and you are inviting sudden disaster in a big way.
On that point, don't forget too that even if your information is perfect in terms of accuracy and resolution (which it cannot be), when you're on the highway with traffic nearby, a situation could start to develop and your ability to respond to it might be limited by the circumstances around you. If you're going to run close to the edge and leave yourself very little time between the point when the alarm bells go off and when the situation gets out of control, then you'll be putting yourself in a position where you might have to make the decision to sacrifice the motor in order to avoid an accident, which would be a shame. In a laboratory sense it's a cool idea to push the absolute limits but your environment in this case will not be especially well controlled. I share the interest in experimentation, but if you're hoping to have the car survive this one, then rather than approaching it with the goal of running right on the line between hot and too-hot and trying to marshal sufficiently precise information to know exactly where that is, I would focus instead on trying to maximize the safety margin by aiming to keep it as cool as possible. The safety of car and passengers/other drivers alike (as well as the success of the trip, of course...) hinges on having maximum time to react to any signs of trouble and aiming to back off well *before* your situation starts edging towards the danger zone.
I usually operate with the assumption that anything outside the range of normal operating temp (i.e. exact middle on the gauge -- let's say up to 100C) is not safe for the motor. With that strategy I've driven mine on trips through the western US under many different weather and terrain conditions, usually with success -- but not always, even using this very conservative approach.
The margin is pretty slim and what's only a little too warm in something else can be way too warm in one of these. I think they're completely solid as long as you keep the temps absolutely in check, but you have almost no headroom to play with. My sense is that 120C as an alarm point could be already way past the point of no return -- I'd be backing way off way before then -- but I'll be interested to hear what happens....