Why would you buy an external monitor that has less dpi than your rmbp, it negates the point of purchasing retina mbp. I would love a retina display, but at 15 inch the display is to small for me to work with, might wait till retina display available for the 27 inch.
I'd still take screen real estate over dpi to a certain point. If they're buying both of these items, hopefully they do intend to take the notebook with them at times.
Calibration can make colors seem brighter and sharper, and can increase contrast which in turn makes most text look crisper. If it's done properly, it can make a huge difference in how the display looks. It still might not be what the OP wants, but is at least worth a try. Anyone willing to spend that much cash on a display should be more than willing to calibrate it properly, in my opinion; that especially goes for people using both displays at the same time. If they're not calibrated, everything will look vastly different on one display than on the other.
It will still look different unless you're taking steps to match them specifically. I realize you probably think I'm being argumentative here, but whenever someone says "calibration" it sounds like hardware values are being tuned to some kind of reference graph. It's not really the case. The colorimeter attempts to measure the output and organize a description of that hardware for the gpu. This can even out weirdness in gamma, which would achieve the effect you mentioned. It cannot actually increase contrast ratios. These software bundles have no direct ability to apply hardware instructions, so what you're seeing is a change in gains there by what is fed to the framebuffer. I could still see the argument of it improving text, but I don't think it will make up for such a huge level of disdain.
If you're trying to match displays, this really requires that kind of feature in the software, otherwise it's too annoying. This means whichever one has a greater range will be compressed to the level of the other. I'm pretty sure the thunderbolt display is technically higher in overall contrast ratio. This has nothing to do with opinions on which is better. I'm simply referring to a ratio between black and white point luminance at greater than 50% brightness. If you go really low, the behavior might go a little more weird. At that point I don't know which would hold up better.
I've tested most of the available calibration/profiling software including i1 profiler, basicolor, coloreyes, spectraview (US version), and some of the older versions of Datacolor's software. I haven't tried all of them in their most recent versions, but I've spent more time messing with this stuff trying to figure out what grants the best shadow detail and greyscale than I wish to remember
. It makes a difference. It's just that given the TB display's slightly restricted options and the complete disdain of the OP, I think he'd be best off not getting stuck with such a display. The colorimeter isn't a bad buy either way. It's worth using on the rMBP.