What about raw power? How does the K2 compare to the K0? In your book and from your post I got the impression the every year after the K0 the engine was detuned a little.
That's correct: the K0 produced a carefully-built 67 HP (or a little more) on every bike built. The K1 let this slip to about 65 in the "old factory" bikes, and a little less in the 'new factory' bikes, from less hand-finishing of the ports and some leaning out of the fuel mixtures to reduce plug fouling by bankers and lawyers who were putting around town in 4th gear during their 1 mile commutes.
Everyone was riding this new machine!
The K2 dropped to about 62 HP, due to a more restrictive set of pipes and an airbox to match, and the new 18 tooth front drive sprocket. The ports were also not very well finished, as this was the era when Honda was really struggling to produce enough of the bikes: the demand was off the charts. The K3 was produced for a shorter time than the K2, but more than 50,000 of them were made: these bikes lost power to loosened cam specs, less-than-great port shapes, and several quality 'slips' that happened as workers were crawling over each other to build them. Meanwhile, the K4, which used fewer parts and had better paint, was being perfected to optimize things at the 'new factory'.
One of those things was lowered compression, to use regular gas in the US where we had a nasty gas shortage going on, and at first this was accomplished with 1mm taller cylinder castings. Later, this distance (0.8mm, typically) was added to the deck surface of the heads because of better molds. The use of the newest-technology injection-molded heads was the centerpiece. This made picture-perfect heads and ports that needed little production 'help', so they came out in great numbers, fast. Some of these appeared on the later K3, but most are one the K4/5/6 and F0/1 engines. They are identified by the lack of the 2 holes in the bottom fins that were used to fixture them for all sorts of work: it was not needed anymore. They are also visible by the taller distance from the edge of the intake valve (or exhaust, for that matter) to the deck surface, where the 0.8mm was added. This increase the chamber volume by about 0.5cc and makes the compression 8.8:1 or so. Naturally, this lost some power, too.
The cam changed for the F0 engine, mostly to meet emissions at 2500 RPM (that was how silly the EPA rules of the day were: idle and 2500 RPM tailpipe sniffs, cars or bikes). The 4-1 headers of the "F" bikes were also implemented to be able to meet the EPA testing rules: it was impossible to measure 4 individual pipes at that time in most emissions stations. Honda was trying hard to comply with EPA demands: yet Kawi, Suzy, and Yamaha were not ever required to meet the specs, and make 2-stroke bikes at the time. This is our government, but that's another topic...
The fule was getting better in the US by 1975, so Honda hazarded a HP increase in the F1 engine. This was done with the PD carbs and an improved 4-1 pipe that acted a little bit like a header (and was 8 lbs. lighter than the F0). All sorts of HP numbers were being advertised then: the F0 was first touted at 72 HP, then at 76 HP. Cycle magazine could only get 66 HP in the F0 test on their dyno, though, so it was thought these new numbers might be "crankshaft ratings" (whatever that meant?) in the American press. Meanwhile, those of us with well-broken in K0/1/2 bikes were kicking the F0 into a corner on the street, and taking on the Z1 successfully. We snickered (and wondered) at the magazines' HP numbers in those days. It came to a head at our shop in 1974, just before I left, when I 'tuned' a K3 for one of our customers who wanted to take a Z1 and a BMW R90 at the weekend drags. The following week, both that Z1 and the R90 were in my shop, asking for "some power increase, because I [we] got beat by an 'old' Honda 750 last week".
I was not directly working on the 750 (in shops) during the reign of the F2/3 and K7/8, as I had just moved to Colorado. It was about 5 years later (mid-1980s) before I got into those models. The larger F2/3 valves (compared to the K7/8) seemed strange to me, and they wear out fast for their guides and stems, but while they last they do breathe better than the earlier engines. The F2/3 also got some extra cam lift (about .3-.4mm on average, as measured - not as advertised), with steep opening ramps to sort of increase the duration without getting the valves too close to the pistons. This led to "snap" marks on the closing sides of many F2/3 cams, which is how their valve trains take such a beating. If the F0/1 cam is used instead, the top end lasts much longer for the trade, while losing a fraction of a HP at 8500 RPM. I have junked more F2/3 cams than all others combined, over these years, because they have a 'snap', or 'crease' where the rocker slaps against the closing side ramp on the intake valves. This is real hard on the rocker foot.
The F2/3 was said to develop (in advertising) as much as 78 HP. In real life, I think they make about 72-74 real HP on the final drive. The bikes, as delivered, came with overly-heavy 630 chains and even O-ring chains, and a very heavy 4.50x17 rear tire, all of which ate a good portion of that power back up in losses. IMO, this probably cost the bikes the street cred they could have had at the time. Lightening up the drive train by using modern chains in 530 size and either 18" rear wheels or tires like the lighter Metzlers can bring back that 'lost' HP.
