Good Morning jrk, vexorg, chris and anyone else whose's finding this thread as interesting as I am.
It's been a number of years since I last stripped and rebuilt an engine entirely and even longer since I built one up where performance was the main consideration. However I have built a number of engines, from small 2 strokes - which are very interesting subjects to extract increased power from - to small 4 strokes - mostly Minis, but I was also involved in a couple of VW Beetle engines going into beach buggies. All these engines (not the 2 strokes of course) had keyed sprockets on both crank and cam shafts. Very accurate valve timing is a critical factor if the full potential of a performance cam is to be realized so you get into things like offset woodruff keys and, at greater expense, vernier adjustable sprockets. Interestingly, perhaps? I had to revisit a couple of the Mini engines which were seeing use in fast road/hillclimb and sprint and decided to change one from a single row timing chain (this was actually the first engine I ever built for extended performance and, in my inexperience, I didn't change the timing chain to the Cooper "S" Duplex type) The other one we just decide to put a new chain on as "insurance". In both cases the new chain and sprockets - always new sprockets with a new chain no matter how good the old sprockets look visually. No matter how good a used sprocket looks there will be wear there which will cause the new chain to wear more quickly - just ask anyone who runs motor cycles, except perhaps a Harley or shaft drive enthusiast? What I found was that when I dropped the new sprockets and chains on using the previously fitted offset keys, the timing "dialled in" exactly the same so I didn't need to use different offsets on the keys.
However we are here considering toothed belt driven assemblies so let's get back to that. This same timing effect is what I've observed on the wee Fiat engines. Yes, having the ability to infinitely vary the cam timing due to the cam sprocket being unkeyed means you can carry out basically the same "dialing in" (using locking tools in this case rather than the degree wheel and DTI guage to do it the "old" way - which, I would argue, would actually be a much more accurate way to do it) But I would argue that once correctly set up and the sprocket locked to the camshaft by tightening it's bolt simply renewing the cam belt, even including a new water pump, would not have any material effect on valve timing. Only a manufacturing intolerance in the placing of the water pump spindle in relation to the casting it runs in would make a difference to the timing and I think that highly unlikely. The tensioner, being on the "slack" side of the drive, would not affect timing and the belts themselves, as long as you buy a "big" name, have no dimensional differences I've been able to detect.
So I think, and have proved to my satisfaction, that AS LONG AS THE CAM SPROCKET IS CORRECTLY ALIGNED ON THE CAMSHAFT BEFORE YOU START you can correctly achieve accurate valve timing without slackening the cam retaining bolt and marking the belt, sprockets and casings (casings just for extra security of reference) before removing the old belt, transferring the markings to the new belt and reassembling.
I would also say though that unless you have a very good understanding of what you are doing I wouldn't do it this way because if something goes wrong during the change - and I can think of a number of possibilities - unless you understand intimately the 4 stroke cycle and how it relates to your engine, you're probably going to end up with seemingly insurmountable problems and possibly, on later model "interference" engines, a ruined engine.
One of the problems I've seen reported, here on the forum and elsewhere - there was a post not that long ago about this very problem, is that, even when using the locking tools, people end up with the timing one tooth out. A recent feature in Car Mechanics Magazine, where they did a feature on changing the cam belt on this very engine, ran into exactly this problem and it took them 2 days to figure out what they were doing wrong. It was obvious that they were following the instructions I posted a link to earlier in this thread which uses the locking tools but does not recommend slackening the cam sprocket bolt. The problem then arises that the locking tools are, of course, installed before slackening the tensioner so the cam belt is in a tensioned state when the crank and cam are locked up with the timing tools. So slackening the tensioner allows the old belt to be removed easily but when you try to fit the new belt you'll only be able to get it on the sprocket teeth in a one tooth retarded position. You'll never be able to pull the belt tight enough with your hands to get it on in the right position, I know, I've tried! What you have to do is remove the little bolt that holds the end of the crank locking tool to the front of the engine and turn the crank slightly anticlockwise - little more than the diameter of the bolt hole should do the trick - until you can slip the belt on. Having done that you'll find you can turn the crank clockwise slightly again to remove any slack and you'll find, once you've adjusted the tensioner, that the retaining bolt will slip back in.
I'm suggesting turning the crank anticlockwise above aren't I and people get very upset about turning crankshafts against their normal DOR (I've worded it carefully this way because some engines don't run clockwise). Yes, under 99% of normal situations DO NOT TURN CRANKSHAFTS AGAINST THEIR NORMAL DIRECTION OF ROTATION. This is because a worn belt/tensioner or chain/tensioer is much more likely to "jump" a tooth or several if the engine is turned backwards and this can be catastrophic for many engines. However, in this unusual situation we are not talking about more than a few degrees of crankshaft rotation and there is no risk of valves and pistons touchingBy the way, if you are in the habit of leaving your car in gear on a hill then engage a low gear compatible with the direction the car is pointing. What do I mean by that? If the car is facing downhill then select first gear and if facing uphill then select reverse. Any resistance to movement being supplied by the engine's compression will still be felt - although this effect is pretty minimal in reality - but if the car does move then the engine will turn DOR and so the risk of displaced timing due to wear in the chain/belt/tensioner/whatever is avoided.
I suspect some people will want to come back to me on all this and I'm awaiting, with great interest and anticipation what now may follow! However I'm off down to see Steven at the wee Honda Garage now to thank him for referring me to the Mobile Tyre guy and for a wee chat - also gives me a nice hour or so of walking along the old railway line, now a walkway, which will get me home mid afternoon just in time for a nice cuppa and a chocolate swiss roll with Mrs J in the garden. Maybe tomorrow before I can reply - should a reply be warranted. Hope you enjoy the above as much as I've enjoyed thinking about it.