Other hidden costs
I like the *idea* of LED or LASER projector "lamps" but I remain to be convinced.
What no-one ever seems to mention about LCD projectors is that the LCD panels themselves also have a finite lifespan. I work in a museum where we have quite a lot of projectors running the "exhibition" and the original-fit LCD projectors had panels with a 4,500 hour rated life, though I've seen documentation that suggests they were sold to the museum as having a 28,000 hour MTBF. I realise that our 7.5 hour a day use isn't typical of a home, but unless you have a normal TV too and only use the projector to watch two or three movies a week, 4,500 hours will creep up on you quite quickly (4 years at 3 hours a day). Looked at another way, if you have to be replacing the LCD panels every second lamp change (lamps are usually between 2,000 and 3,000 hours) then the cost of the lamps becomes a very minor issue. I wonder whether being illuminated by LED or LASER would increase this lifespan?
In practice it's not just the panels but also the colour filters - you will start to notice lower contrast, even with a new lamp, and maybe colour blotching (LCD failure - the blotches are often blue) or a colour cast (filter failure - yellow). We were quoted 5,000 EUR for a replacement "optical block" (3 LCDs, 3 filters and the prism). This was ridiculous when you consider that the projectors were also beginning to fail in other ways (e.g. PSUs not booting back up after a power down - 1,000 EUR or £3 of capacitors to DIY) . Instead we switched mainly to twin-lamp single-chip DLP projectors, the oldest of which are now around 11,000 hours and are (almost) as good as new (really must get around to that 10,000 hour service!). These particular models have lamps that last twice as long as the lamps in the old LCDs and cost significantly less. Twin lamps are a boon for us giving both security (if one fails, the projector continues to run albeit at reduced brightness) and flexibility (one of the projector models has effectively four output modes - both lamps or a single lamp, high power or low power - allowing us to choose brighness and lamp change interval).
DLP does have its problems, with fringing probably being the most annoying, but if that's an issue then all you need to do is to look at three-chip DLP or possibly (there are still issues, but they're not quite as bad) one of the DLP models with colour wheels with extra colours or which run faster.
A technology no-one has mentioned yet is LCoS. My warning here is that our five LCoS projectors have not lived up to the hype. When new, the picture was excellent but despite being a sort of hybrid of DLP and LCD with the intention of taking the best from both, the panels fade in exactly the same way as LCD, and at about the same age. On top of that the models we have seem to have "open" light paths and dust gets onto the panels quite easily.
As for the subject of the article I have to agree with others here. Half the price and twice the brightness might make it a good choice, but 1,600 lumen isn't really suitable for any room where you have any amount of stray light, unless the image is small enough that you'd be better off spending money on a nice LCD or plasma telly. At the museum, our lowest output projectors are nominally 2,000 lumen (twin lamp units in single lamp modes) which works, but it works mainly because the projected image is no more than 36" horizontally.
Finally, and I realise again that our use in a museum isn't terribly comparable to home use, it's worth considering networking. All our projectors have network sockets. Most will email a preset address when there's a problem (for example, they will email when the lamps have run for more than a certain number of hours) and all can be started and stopped by network messages (PJLINK). This is a prerequisite for us (who wants to go around manually starting 30-odd projectors?) but could also be useful in a "connected home".