A couple thoughts on the foregoing discussions.Cam Considerations
Regarding a cam for the Knucklebolt project, if someone wants the torque/HP balance to favor torque rather than HP (higher rpm operation) then you'd want cams with shorter duration (225-235 deg at 0.050" lift) and relatively tight centerline. The torque will be high, build quickly, then be done. Lift really shouldn't be much of a consideration here since any flathead cam considered for the project should lift the valve to it's flow potential (usually ~ 20-25% of the valve dia - this is why all the K model cams, including KR, only lift ~ 0.380-0.395"). If you want an engine that is well behaved, yet makes some HP (has to rev to make HP - and of course "rev" is a relative term) you'd want cams that had at least 240 deg duration at 0.050" (KHK cams). Larger engines, e.g., 74-84 cu in can use more cam, so the KHK lobe in a UL would not be "big", as large flatheads can accommodate lots of cam. Compression Considerations
As has been noted, for an Otto cycle engine with all other things being equal, the higher the compression ratio the higher the efficiency and the higher the resulting torque/power. And of course since power is heat, the more power the more cooling required. http://victorylibrary.com/tech/SV-compression-c.htm
As we all know and as Dick already noted, too much of a good thing isn't always the best thing, i.e., Frankie, with CR of > 9:1, when hot and under heavy load detonates. Frankie is therefore not a bike to lug around town, and operating it in such a way that makes it detonate consistently would bring it's life to a premature end. I'm certain Dick will confirm that he makes good use of the gearbox and keeps the engine running free the majority of the time to circumvent detonation. The UL, being a considerably larger engine yet, has commensurately larger cooling demands and in the same configuration would be that much less forgiving.
Detonation, the product of heat and time, is not tolerable and if it occurs on a constant basis while riding, leads to only one place - a melt down. Thus, when considering the possible compression ratio for a flathead project, the owner must decide how he wants to use his bike. Is he going to lug it around town and roll it on at full temp in high gear at low rpm (these are the worst conditions for inducing detonation - lots of heat and time), or are you going to gear it a bit lower, and operate the engine at higher rpm in a more free-running state (opposite of lugging), where it will tend to run cooler, and only roll it on for brief bursts when the engine is operating at the torque peak or higher (this reduces the potential for detonation since higher engine speed shortens the combustion time in which detonation occurs).
That said, and in light of a given riding style/desired use, what sort of compression ratio can a flathead tolerate? For all out performance where the engine is run at WOT for sustained periods of time (KRs), racers settled somewhere in the range of ~ 5-6:1, which allowed the engines to survive without melting down. For drag racing or similar short excursions of 10-15 sec, where you start with a cold or lukewarm engine, this situation is much more forgiving and can tolerate higher compression (if you are willing to accept the potential for lost flow to gain compression) without melting down. But for a large displacement air cooled flathead, running on gas, at full operating temperature, with over 7:1 CR, I think that operating it at WOT for any length of time would result in overheating, detonation, and in turn engine melt down. For street use I personally wouldn't be looking at anything over 7:1. It is my opinion that ratios higher than 7:1, come at the expense of flow, and excessive heat generation. Airflow Considerations
Relative to their OHV counterparts, flatheads by their very nature suffer from a compromised state of breathing, because ~ 40% of the circumference of a flathead valve is masked by the cylinder head. KR heads are generously relieved around the IN valve circumference to decrease such valve masking and increase air flow. So if an owner starts filling in the combustion chamber around the valve circumference and/or significantly decreases the height of the transfer port (by milling) to reduce combustion chamber volume, this may compromise flow to an extent that any benefit gained by increasing compression is more than offset by lost flow - thus the conundrum that Dick alluded to. Pop-up pistons provide a means of keeping the transfer slot proportioned appropriately while preserving compression ratio. However, I believe in severe use applications, such as the KR, the piston crown was modified by essentially removing the top of the piston on the valve side to shield it from the propagating flame front when it was fully exposed at TDC (of course this was another compression loss, but one that was lived with for the duration of the KR's reign). The cast KR pistons were typically fit at 0.007", which is really loose for a cast piston. Yet I think it speaks to the sort of heat encountered in the application, and likely at full operating temperature the piston fit was satisfactory.
The underlying tenet running through the majority of the above discussion is that flatheads are heat machines, that compression and power mean more heat, that more heat means increased potential for detonation, or need to retard ignition timing (which in turn adds more heat), all of which can lead to unhappy engines. Heat in a flathead is not only undesirable, it is by far our worst enemy. However, if we plan and execute our projects carefully, the results can be reliable and satisfying.
Hopefully the above discussion provides a little food for thought as builders decide how they intend to operate their engine and which performance features might allow the engine to live happily ever after.