I spent an awful lot of time a year ago trying to make both the Enterprise and the Kelvin work at TOS scales and couldn't get it to work. The real problem is the bridge: from the top of the dome to the base of the platform underneath it wouldn't cover the height of the saucer rim; not even close. And the bridge dome is almost certainly a meter or so taller than the actual deck height, which necessitates a two-deck saucer any way you look at it.
Ah, but the interior view shows that the bridge wouldn't start from the top of the dome. Rather, even its outer rim (and not just the middle pit) would be sunken a bit into the second tier of the superstructure.
Which makes virtually no difference. The bridge deck would have to extend all the way to the bottom of the second tier in order to fit the ~300 meter scale; this alone would be consistent with a single-deck saucer but would be inconsistent with a bridge that doesn't completely fill the dome and WILDLY inconsistent with the shape and size of the viewscreen.
The scale you're using here would put the Kelvin in the neighborhood of 400 meters; that still wouldn't fit the interiors or the shuttlebay, nor would it give you a one-deck saucer (actually, closer to one and a half).
If that guy were a normal-height fella, the ship could be as small as the Saladin.
Kelvin doesn't have the same proportions as Saladin; it being the same
size (similar diameter though much thinner saucer) it would still be around 400 meters long. Again: the only way to fit a one-deck saucer is to have the bridge dome and second tier BOTH fit into the first deck; that would give you your 270 meters, but it would leave you with a Kelvin that is smaller in every measure--by volume, by mass, by size, by diameter and thickness of the saucer--as NX-01... it would be considerably SMALLER than the Saladin despite having the same length.
As you of all people should know, length and size are two different things. Kelvin just isn't that large of a ship even at 457 meters, and the problem is no other size can be made consistent with the interiors we saw.
At the end of the day, you're left with the fact that a 270 meter vessel wasn't in the film, neither was a 300 meter one. The ship that we were presented with was 457 meters long; you're better off finding ways to make your analysis consistent with THAT instead of trying to shoehorn visual cues to fit your own arbitrary assumptions.
It seems pretty clear that the ship was designed to be a certain size
Which is irrelevant since the CANON size is 457 meters long. Visual cues need to be interpreted to be consistent with its established size, not the other way around.
Also, the richter mesh is different from the actual ship (relevantly, in this case) by the shape of the bridge dome and the second tier superstructure.
You can see here that they are approximately the same height, and the FILM's model has a bridge with a small platform directly underneath it; there's no reason to assume it's recessed into the second tier, as based on this image the bridge is CLEARLY meant to fit into a single deck of the bridge dome.
Perhaps you are confusing Tobias Richter's scaling intentions with those of ILM? Have you considered that possibility?
That's "the glass is half-empty talk". 
