Retina Display is unhealthy, but it's unrelated to the resolution. It's the fact, that it's offered only as glossy display, no matte option.
Queensland University of Technology page on "High gloss computer screens"
This web page contains health and safety considerations for Macintosh
– Apple ‘glass’ or high gloss monitor screens.
Reflections and glare on high gloss monitor screens and their relation
to the angle of the monitor screen, could cause the operator to adopt
awkward postures when viewing the monitor screen and using related
equipment. These reflections on the screen can be from internal and
external sources such as the overhead lighting and/or position of
Awkward postures adopted by the operator may in turn lead to an injury.
Glare is also mentioned by American Optometric Association in their description of "Computer Vision Syndrome"
symptoms may be caused by:
- poor lighting
- glare on the computer screen
- improper viewing distances
- poor seating posture
- uncorrected vision problems
- a combination of these factors
Note, that they do not mention resolution as one of the factors.
Popular guidelines talk about ".28mm dot pitch or less", which is equivalent to 90+ dpi. For comparison, really low end laptop displays: 1280×720 at 13.3" is .23mm dot pitch (116 dpi), 1336×768 at 15" is .25mm dot pitch (103dpi).
On the other hand, these guidelines suggest that bigger screen is better, MBP Retinas are offered only as 13" and 15" sizes. That of course isn't a factor if you're comparing eye strain with other similarly sized laptops, but implies that much bigger external displays are better.
I have not seen a single scientific study, which would suggest any particular resolution threshold, much less study suggesting going below .28mm would reduce eye-strain further. Which of course doesn't rule out that it might be the case.
As for the question why would Retina appear to be "so much better" than 135dpi screen...
Although there haven't been any serious studies for Retina in particular, in the past it has been shown that there is kind of "placebo effect" when it comes to screens or content in with higher resolution. For example 2008 survey showed, that:
survey by the Leichtman Research Group (LRG) shows that 18 percent of
HDTV owners think they're watching high-definition shows, when in fact
they're viewing standard definition programming. The findings are
based on a telephone survey of 1302 U.S. households. This is LRG's
sixth annual study on the topic.
Given the general consumer confusion surrounding HDTV-all those
mind-numbing specs like 1080p and HDMI aren't easy to grasp-it's no
surprise that many buyers are still clueless. Standard content that's
stretched to fill the entire screen may look funny, but at least the
Retina vs standard res: There have been multiple a bit tongue-in-cheek "studies", where people would be tricked into believing they are using iPad 3 (with Retina display), while in fact they were using iPad 2. Still, they would describe image as much more crisp and higher res than on iPad 2 (sic!). For example one of such a experiments was conducted by Gizmondo.