I was apparently wrong.
I just looked at Thomson's installation instruction, and it specifically does say to grease the threads and heads.

But I am quite sure I have seen it both ways, where it is specifically warned to not lube the threads and heads.
It depends upon what assumptions the designer was making at the time they specified the proper torque. But apparently grease on treads and heads is not such a bad thing.
I never like to put things together dry, but I prefer anti-seize or loctite, just because greased threads have tended to loosen under stress and vibration.

Most manufacturers for bicycle components give torque specs for greased bolts. Easton specifically says to grease the threaded bore and not the bolt. Occasionally, loctite will be specified (i.e. rotor bolts). Really, if you're in doubt, especially using fancy carbon components, you should read the user manual.

As for the specific fork: if the steertube is compressing enough at the top that there is a visible difference in how far you're tightening the stem, I second the idea of a better expander wedge to ensure you don't pinch the steerer.

Trust me, I am a compulsive manual reader and don't put peanut butter on a slice of bread without checking the jar for recommended thickness

The manufacturer (Deda Elementi in this case) recommends all of 8nm which seems a lot to me. But the uneven gap occurs already around 5-6nm. I am actually more worried about creating tension in the stem (which I am afraid might lead to cracking) than pinching the steerer.

I have noticed Deda puts some high torque values on its stem bolts for the steerer tube clamp. There is no way you need 8 Nm, any quality stem and fork combo should be fine at 5 Nm. I never used a torque wrench until recently, and I learned that my "feel" torque for stem bolts on the steerer was about 4.5 Nm and I have never had problems.

As has been covered many times before on this forum, the torque values stamped on stems, bars and other parts are NOT recommended torque, they are the MAX torque. The REQUIRED torque is the amount that is necessary to firmly hold the part in place. For example, most 4-bolt stem plates only require 4-4.5 Nm of torque to firmly hold bars in place, not the 5 Nm or 8 Nm that u often see stamped as the MAX torque value.
EM3

_________________
________________

My 2011 Cervelo R5

Also note that if in doubt, do not reach max torque with a greased thread. You will reach rated torque at a stress value that is less than if you had greased. The same torque is still there to prevent loosening, but you have reduced stress.

That is a revelation to me sir. I thought it was recommended. You have possibly saved my day although I am now forced to use judgement rather than the manual which does not sit well with me

some companies post recommendations, some post max values, and some like Deda and others post a maximum torque that type of bolt should use if threaded into a steel object (unlike the stem. eg the torque listed has nothing to do with the application).
thomson states to grease the bolt threads AND the threaded holes in the stem before assembly.
thomson states to only grease the threads of the bolts on their seatposts, and explicitly states NO grease on the bolt head or washer. and no anti-sieze.
on most 4bolt stems with m5 bolts, i tighten the faceplate to 4nm. zipp says 4nm, specialized says 4.5nm, i push as hard as i can on the bars to see if they will slip when it is with a product im not familiar with to see if 4nm does the job.

This is absolutely correct, and that is exactly the purpose of carbon assembly paste; to reduce the torque required to get the part to hold properly, guaranteeing you will not even approach the maximum torque of the part in question.

As for lube - the interference face of any bolt should be dry, but the threads should have anti-seize paste or grease on them. This is not so much a lubrication issue as it is one of dissimilar metals in contact with one another. The grease or paste will ensure that you can remove the bolt much later without breaking it. Automobile wheel bots rely on a slight deformation of the soft bolt shoulder when torqued properly, and should never be lubricated. Other bolts are selected for head type, etc., based on their proposed usage and torque requirements. The variance in stress when torqued to spec WITH lube on the heads from one bolt to another can be quite large, due to the variance in the type of lube used and application.

Bottom line - use a torque wrench, do not lube the heads of the bolts, DO use carbon assembly paste, and torque all bolts progressively to the same torque, which will be considerably less than the max. torque specified on the part if you use the assembly paste.