There are (depending on how your count them) four Steinway D scalings going back to 1876 (but I know Steinway designed and made concert grands before this time--including a 1859 overstrung scaling visible in Chapter 1 of Alfred Dolge's "Pianos and their makers"). The original scaling (if my sources are correct) debuted at the Centennial Exhibition in Philadelphia, the first World's Fair in the USA which commemorated the 100th anniversary of the signing of the Declaration of Independence.
This first design, which was usually called the Steinway 4 or Steinway Concert Grand in their marketing literature was 270cm. In 1878 all models were redesigned and rebranded, so the Steinway 4 became the D, the Steinway 3 became the C, etc. In the main, these scalings changed very little with the 1878 rename, but minor improvements outside of the original design were constantly being incorporated into Steinway's pianos during these years. Until 1884, all Model Ds were 270cm, and usually listed by contemporary salespeople as "Centennial D-270s" which are extremely rare--less than 500 remaining examples of the D at this size. Of all of these pre-1884 instruments, none are really identical in design, unlike the future "D" models which are quite consistent. Also, it is not clear if most or all of the pre-1884 instruments were 85-key instead of 88-key. Some sources say nothing was 88-key until 1884, but other sources disagree, which makes me cautious of anyone who lists an antique Steinway with 88-keys and says it pre-dates 1884 or even predates the 1876 Exhibition, but I really can't say on any such cases without consulting Steinway's archives. I also know these are the American dates, I'm not sure how much delay between design innovations being made in Hamburg (after 1880) and New York there might have been.
In 1884, after a total redesign of the D (including the addition of the capo d'astro bar and lots of internal scale changes), the new rim was 272cm, which was the rim size until at least 1896, but the rim change doesn't appear to change the internal string length just the width of the rim. I'm not sure when exactly Steinway went in their marketing literature from this new 1884 "D" to the current "D-274" but there aren't many other changes to the instrument--as far as patents go--since then (Patent #1826848 for a faster repeating action, Patent #2051633 for a more resonant soundboard, and Patent #3091149 for a tighter pinblock). I'm not sure, since I don't have a copy of Kehl and Kirkland's "The Official Guide to Steinway Pianos" which could have more detail than I can easily find in my sources. It's strange as a lot more documentation exists online and in my source books on the different iterations and scalings of the A, B, and C than there seems to be on the might "D" itself. I'm rather disappointed that Cyril Ehrlich only talks about square pianos and Steinway's marketing rather than their 19th century grands, and David Rowland only mentions Rubinstein's America tour for Steinway in 1872-1873, which I doubt was on a square piano, instead focusing on overall sales of American pianos in the European market and he doesn't even discuss 85 vs 88 key range (which is unusual considering his thorough documentation of historical piano compass from 1750-1850).
To the question of why 270 instead of 274, my guess is that either PTQ left it the older string/case length to prevent licensing problems with the fact that only with PTQ 5, did Steinway authorize their marque to be used in PTQ and the D4 became the "Steinway D" and this is an artifact that time in PTQ development. Or, case size and string length doesn't necessarily overlap as easily as we presume that they do: e.g. since strings sit at a slight angle in most pianos for overstrining, they may be noticeably different in length when compared to the case. I can't easily measure this on multiple pianos, but it's possible that--especially as the 270 to 272 transition was only an increase in rim size not string length--the change from 272 to 274 was more cosmetic than structural as far as string placement, length, or density and 270 is better representative of the D than the rim measurement. It's further possible that the string composition (material and gauge which sadly we can't change yet) mean that a more authentic tone is created by the PTQ engine at 270 default length than it is at 274 length, since sound quality should overrule arbitrary specs.
While it's possible that this is to create more of a parlor feeling in tone, I suspect PTQ would have just modeled a C-227 semi-concert grand (which was Steinway's original parlor offering) instead to achieve that (especially as they're still readily in production to fill the gap between the D and B).
Only Philippe or another Modartt designer could tell us the exact reason.
Link to Dolge for the 1859 illustration: https://archive.org/details/pianostheir...2/mode/2up
Spotify:
https://open.spotify.com/artist/2xHiPcCsm29R12HX4eXd4JPianoteq Studio & Organteq
Casio GP300 & Custom organ console