Main things to look out for?

[MaxD]

Max, one of my books says the exhaust gas temperatures can approach 1700 degF.  The point where I was measuring was the part of collar around the bolt that is on the far side from the exhaust pipe.  That's about 1.5cm from the pipe, which is benefiting from the cylinder fins and general heat spreading.  I did not try the inside that is about 0.5cm from the pipe because I did not want to ruin my temp probe, which is intended for electronic work and won't stand up to what might be 1000degF (nor will my hands!).   

We don't seem to have the engineering data here to make the best call.  Suzuki does know, but their decision to use cheap apparently zinc plated steel bolts is apparently flawed.  If I were making this call right now myself with this limited data, seeking the most reliable and least risky solution, I would say stainless steel bolts with anti-seize torqued to 60% of recommended dry torque, loosened and  re-torqued periodically, and replaced every few years to eliminate corrosion that still occurs even with the anti-seize.  Of course, I'm a complete amateur and just about everybody here knows more about bike maintenance than I do.  If the 271 Loctite is in fact holding on the far side of the threads from the pipes, then the studs would seem quite safe and more convenient. 

[MaxD]

I have a professional update on the Loctite for use with the header studs which is not definitive, but which is still fairly supportive of using the studs with Loctite. 

I went for advice on this to tech support at Loctite, specifically to support engineer Sylvain Lebel (United States support, sylvain.lebel@henkel.com, 800-562-8483).  He has worked there 16 years and has had many calls on the question of using Loctite in these high temp applications.  His bottom line summary was that while they do not specifically recommend any version of Loctite for header stud application, that quite a few people including pros are using it for that, and he is not getting any complaints of the studs moving with nut tightening and stripping out aluminum threads.  The most common grade used is the widely available Loctite 272.  Henkel/Loctite does not warranty Loctite, and users are on their own for all applications. 

Further details are:
1. Well above spec range Loctites will "burn" and loose effectiveness.  They have not tested for this to report specific consequences.  Since they don't warranty, they are not that worried about nailing down the exact consequences. 
2.  Grip strength degrades fairly linearly with temperature.  They spec the range at the point where grip strength is about 50% of room temp. 
3. Loctite 272 will have both temperature and age related weakening in this application.  After 1000 hours at the high end of the spec range of 272, which he says is actually 450F, the room temp strength will be down to 60% of original.  At that use time while at the high end of the temp range, strength will be down to about 30% of room temp original.  Sylvain is not aware of the number of cycles being a problem--its a decline based on total time at high temps.  When cycled it does not melt and turn liquid, but it does soften into a plasma like state.  Apparently it can then re-harden without damage, but this is an assumption (no warranty on cycle count effect).   
4. He warns against trying to remove any high strength version at room temp--it can tear aluminum threads.  It must be heated to be removed. 
5. There are recent higher temp versions of Loctite available.  2422 is a medium strength 650F version (50% of strength at 650F).  2620 is a 650F high strength like 272.  These are not as available as 272, but can be had from industrial suppliers.  How much they are being used yet for header studs is not known, as they have only been out about 2 years. 
6.  Loctite in its just applied state does act as a thread lubricant, so reducing torque to about 60% of standard dry torque is recommended.  The issue of just how much studs should be tightened in this application (finger tight, a little more, or 50-60% of bolt spec) is still not clear.   

So, that gives some support for studs with available Loctite 272.  It is not warranteed or recommended, but empirically it seems to be doing OK in the field over a number of years.   It may be slipping and tightening the stud on nut torquing if the Loctite is burned, but stripping as a result is not known to Loctite to be a common occurrence.  Therefore, it might be safer to use studs with Loctite than bolts with anti-seize or without, because it is avoiding the need to re-torque bolts periodically and thus subject the threads to repeated torque hits.

That's as much info as I know how to get on trying to correct this header bolt corrosion design problem.  It seems there is no known highly reliable cure, but we have taken a step in the right direction with either stainless steel bolts, or with stainless steel studs and Loctite 272.  In either case we have to check and possibly re-torque either bolts or stud nuts fairly often.  We also have to replace bolts when they get significant corrosion.  That is presumably what Suzuki was expecting to be done, it was just that their zinc plated bolts probably have to be replaced more often than stainless steel bolts. 

[MaxD]

Update:  See post "Suzuki knew what they were doing after all" for the full story.  It is almost certain Suzuki was deliberately using zinc plated header bolts as a "sacrificial anode" to prevent galvanic corrosion of the aluminum threads.  Stainless in aluminum will force the aluminum to become sacrificial and corrode.  This might be fine for stainless studs, especially if they are well sealed.  For stainless bolts that are periodically tightened and sometimes replaced, the aluminum threads are corroding and the time will probably come when they won't hold a new bolt.