It's about a year and a half since the firs post on this article. Let's revisit and review what has happened and there's anything new worth mentioning!
First off, as far as Canon goes, the 60d and the 600d came out. Basically all the same sensor, digic of the 7d and 550d. And the predicted articulating/flip LCD was a feature of both. A great boon to video. Even better is that the 600d also allows for better audio control. For me, the one remaining weakness is a decently usable AF in video. Otherwise, the 60d and the 600d isi as close to "almost perfect" as it can be. As far as photography goes, it's already very good. In fact, almost excellent. It is hard to think of any more improvements in these models as far as stills goes.
Of course, one will say better DR (Dynamic Range), and IQ. But I think, to go that route will require a next gen tech on sensor design. Current/existing tech will never exceed it. If any at all the best we can hope for that will be a big boon is video AF. I don't know if Canon will deliver it on the 70d and the 650d. Likely, it will be an improvement. Sony and Panasonic (esp Panny) has shown a really wicked fast AF on CD (Contrast Detect). Canon and Nikon cannot lag on this area. It will be a sticking point later on. At least Nikon is getting better. Perhaps, this is due to benefiting from it's tie-up with Sony.
In short, I expect the improvements in the video areas as many needs to catch up on video AF speeds. The flip screen will be a staple and no doubt, the 7D mk-2 will have that too. And a 7d mk2 will likely be due end of the year or early next year. The 7d will likely be on a 2-year update cycle. The xxD on a 1.5 year cycle. But if we are to take a peek at what is coming, likely, we need to look at the xxxD bodies, as they are in a 12-month cycle.
Of course, without companies saying, so, the future really is not in DSLR but on MILCs/EVIL/LIVE or pentaprism-mirrorfree bodies. So if there is any breakthrough, it will be there. Only with a mirrorless-pentaprism free body can we break free of the 12fps limitations. With no mirror to hold us back, we can go 15-120fps easy. That would require a global shutter on the sensor and a faster ADC-DAC or processor chips to handle that load. That will take time. But I am digressing here. In any case, the "death" of DSLR is still 7-12 years away. No need to worry about that. But once Canon brings out their MILC model, the ball will move faster. As it is, it is still a big question WHEN they will come out with their own. For now, DSLR is still viable for many applications.
So, if one has a 40d or lower, the 60d is a great upgrade. Again a 2 generation (3-4 year) wait. If one has a 400d or 450d or lower, the 600d is a big upgrade too. But I must emphasize, if you don't need it, no rush to upgrade. If a 350d does what you want, you should not be pressured into getting a new body.
As far as the tips goes, they still hold well. The basic rules still apply. If one has invested in lenses instead, it will still serve you better and longer. Even with the coming of mk-2 lenses (e.g. 70-200 f2.8L IS mk-2, etc), their mk-1 counterparts are still good. Just because the mk-2 came out, it does not mean the mk-1 are coke bottles. The improvements are marginal and significant only in the areas of extreme use conditions (e.g. flare control). It may or may not be useful to many, but as an owner of the 70-200 f2.8L IS mk-1, I am not worried that my mk-1 is obsolete or not as good. For me, there is no compelling reason to upgrade, especially if I have to shell out P45-50k to upgrade!
This illustrates again, that lenses hold their value well, and as far as DSLRs goes, one's money should be put in building a system around lenses rather than bodies. Bodies are the disposal items. Lenses are the ones that drive the image. An excellent lens on an old body will give more benefits than an expensive body on a mediocre or poor lens. There are exceptions of course. But in general, the guide/rule is pretty sound and stable.