Bruce B. (`83 Magna) experienced one failed cam and two rockers at around 15k. Greg S. (`83 Magna) experienced a worn rocker at 18k... repair quote, $350-400. Geoffrey Y. (`85 Magna) noted a scored rear exhaust cam at 18k. Shawn H.(`84 Magna) replaced his cams and chain tensioners at under 29k... cams cost $70 each, tensioners $45 each, from Banzai. Keith M. (`83 Magna) replaced his cams, rockers, etc. at under 40k. Ray T. (`84 Sabre) replaced his cams, rockers, etc. at under 40k. Gaylon G. (`86 Magna) replaced his cams and rockers at 50k. Despite these cam failures, Rob T. has 107k on his `85 V65 Sabre, and reports that the original cams are fine, and have only light pitting if anything.
According to Art Reitsma, some of the earlier cams were so soft, they would not even register on a Rockwell hardness tester. At one point, Honda apparently upgraded the quality of their cams. However, it is still unclear to me exactly when, and in what models, this took place. Some say that `83 V65s have inherent cam problems, but the later ones are fine. Some say that only the early V45s had inherent cam problems, and all V65s are fine. Some feel that all V4s before `86 are subject to cam failure, and that truly adequate replacement cams have only appeared more recently. There have been random cases of failed cams in V45s and V65s up to at least 1986, no clear pattern. Regardless, the cams offered now by Honda are apparently improved versions, with a more durable surface, and extra lubrication grooves on the journals, as noted by Shawn H.
Cam failures seemed to originate in Europe, where higher speeds, and thus higher rpms are more common. Due to this observation, it has become common belief that high rpm usage is what hurts the cams. In support of this, most enthusiasts who admit that they ride hard, have had to replace their cams and rockers (along with clutches, transmission parts, etc.). The many enthusiasts who claim to ride in a sane fashion report no apparent cam problems. With this in mind, many enthusiasts are satisfied believing that their V4 cams will remain intact as long as they don't beat the living daylights out of the engine.
However, a Honda rep recently told Ray T. that prolonged idling will hurt the cams too. The rep explained that during idle, the oil pressure to the cams is inadequate, and in addition, idling allows the cams to get very hot. These two factors combined spell trouble for "weak" cams. Some knowledgeable enthusiasts specifically recommend keeping revs UP in order to provide adequate oil pressure to the cams. They recommend raising the idle adjustment, and also recommend cruising at a slightly higher rpm than what might normally seem fit. In addition, some swap in lower temperature radiator fan stat switches to keep engine temperatures lower at idle.
It appears that in both idling and high rpm situations, the "soft" cams are not getting enough oil to keep them intact. Also, it is a fact that the cams/rockers receive unfiltered oil of a relatively low pressure due to the stock oil system design. To improve the situation, various aftermarket oil modification kits have been produced to provide filtered oil of greater volume/pressure to the cams. (see OIL KIT - MODS)
You can purchase new OEM cams from Honda or Banzai, or have your used cams reground and hardfaced. Outfits such as Web Cams, MegaCycle, Shadbolt, etc., offer such a service. Keep in mind that the regrinding and hardfacing of used cams can be more expensive (about $450/set) than a set of new OEM cams ($300/set from Banzai). The article by Art Reitsma about Shadbolt suggests otherwise however. Regardless, you may have a better chance of receiving a true high quality "hard" surface when specifically having your cams hardfaced as opposed to buying new stock OEM cams, since mass produced parts may vary in quality due to wider error margins. Plus, the aftermarket outfits provide slightly wilder grinds for those who might enjoy a bit more top end.
Don't forget the rockers (cam followers). You wouldn't want to install new cams along with damaged pitted rockers. Web Cams specifically recommends using new Honda OEM rockers with cam replacement.
Another hardening process that can be applied to the cams/rockers is "deep cryogenic tempering"... parts are frozen to minus 300F degrees. Such a process costs about $12.50 per cam, $1.50 per rocker arm, plus shipping. For more information, contact 300 Below Motorsports 930 W 23rd St., Suite 19, Tempe, AZ 85282.

Web Cams also offers similar grinds to the Megacycle, as well as Dodge promotions.

Since most of the above specs have been calculated, they may be inaccurate. In addition, Ray T. and Shawn H. have noticed that their stock Honda cams have asymmetrical lobes. The fact that the above calculations are based on symmetrical lobes is another reason why the above figures are probably inaccurate. I chose to provide the figures anyway, as they serve as a quick comparison guide. The only way to truly figure out the exact profiles, is to fully degree and plot the cams as installed.

Art Reitsma refers to the asymmetrical pattern of the Honda cam lobes as having a "backside hump, to allow the valve to stay open longer, after the intake flow has been established, allowing more in for more power". According to an Ultradyne Cam catalog, the benefits of asymmetrical cam lobes are as follows: the intake ramp is steep and quick, in order to open the valve fast, thus allowing a later opening point while retaining similar "area under the curve" compared to a regular pattern with an earlier opening point. A later intake opening point will help reduce reversion and overlap, and thus improve low end performance and economy. On the closing side, the ramp is less steep, and thus "humped", in order to provide a gentle valve closing and thus prevent valve bounce and float. I'll add that with the 10,000 rpm range of the V65, anything to help prevent valve bounce/float is necessary. If switching to aftermarket cams, stick with the asymmetrical type.
According to the Ultradyne catalog, the way to degree asymmetrical cams (find the lobe centerline), is to use the "50 down" method. In other words, instead of taking measurements at .050" or 1mm" of lift as usual, take the measurements at a lift that is about .050" below the maximum lift of the lobe. It does not need to be exactly .050" below, but an easily read lift nearby. Then, as usual, grab a reading from the front ramp, then one from the rear ramp, find the difference between the two, divide by 2, and that's your centerline.
Concerning setting the position of the cams when degreeing, you will need to slot the cam sprocket attachment holes, according to Web Cams. They suggest that Falicon (813-797-2468) may provide this service. Ray T. suggests that NRC (Niagra Race Crafters) of Toronto, or APE of California may provide a sprocket slotting service as well. Dale Walker probably knows something about this too.
An article by Art Reitsma relates that Geoff Bardal of Shadbolt Cams has witnessed stock intake cams as having a duration of 227 degrees and stock exhaust cams of 224 degrees. Geoff came across a "rare" set with 227 int, and 229 exh. None of these figures match the Honda V65 shop manual specs. Perhaps Geoff is referring to V45 cams? Also, it is not unusual for service manuals to be inaccurate about such things. In addition, it's possible that cam profiles were changed a bit over the years. If you really want to know for sure what's in your bike, or check effective chain stretch, grab your degree wheel and do some measuring. Has anyone out there degreed their stock cams? Results?
The stock valve timing is fine, needless to say. Plenty of low end power with peak horsepower occurring roughly 500 rpm before redline. Perfect. No need for a wilder cam grind, at least for street use. If anything, riders who spend a lot of time touring with a heavy load can probably benefit by advancing the stock cams, in order to increase low/mid range power.
Though on paper, the Sabre's stock cam timing is a bit wilder than the Magnas, the real world difference is not that great. Gaylon G. swapped a set of Sabre cams into his Magna, and stated that he could not feel much of a difference in performance. Also, if you study the Sabre and Magna magazine test articles, you'll note no significant difference in performance between the two models. (see DIFFERENCES B/W MAGNA & SABRE)