First off, I should reiterate that the internal layouts were not done by Masaro Okazaki, the creator of the site, but by Allen Rolfes, though Allen credits Masaro with a great deal of helpful suggestions. It's all in his notes.
Next, I feel it's fair to be more specific about my criticisms.
When I said "The [deuterium] volumes are demonstrably wrong in several cases," I was not necessarily speaking of the Constitution class, of which 1701 is a member. Any of the M/AM powered vessels have so many unknowns that well formed criticisms are hard to construct: Too many assumptions are needed to make conclusions. However, there are a few things that can be said about the deuterium storage of Mr. Rolfes Constitution design.
In his notes on the Constitution class, Rolfes speaks of 924 metric tonnes of deuterium for 6 months of operation. That deuterium is spread between three storage tanks: one in the dorsal Engineering Section and two Starboard and Port aft quarters of the saucer section, deck 7.
Calculating with 30 days per month, that's 59.4 grams of deuterium per second. Used in a simple D+D fusion cycle, where 0.0973% of the mass is turned into energy, that's 5.2 terawatts (10^12 watts) of power available for conversion, on average. (Remember, 1 terawatt = 1 terajoule per second.) There would be some conversion inefficiency, but even using a terrible thermal conversion efficiency of a third, that's still well over a terawatt of useful power. For comparison, the US average power use in 2004 averaged around 3.3 terawatts.
The above calculation is the lower bound. If you use a catalyzed D+D fusion cycle, where 0.384% is turned into energy, you get over 20 terawatts of convertible power. And if you go full M/AM, it's 5,340 terawatts. It's clear from the design and the notes, however, that M/AM reactor is not meant as the main power provider. Even so, since we can't know what percentage of power is taken from the warp reactor, we can use the M/AM number as an upper bound.
Again, there's nothing wrong with any of these numbers being perfectly workable for a 23rd century starship. Indeed, it doesn't seem likely she needs a terawatt to run her systems at idle. Far more likely, she needs huge bursts of power during combat and emergency maneuvers. So I have to conclude there's nothing at all wrong with the amount of deuterium stored in Rolfes Constitution design.
It's a handful of the pure fusion designs that have me shaking my head a little. The clearest example is the Romulan Clavicle cruiser, from the Earth Romular war. He gave her 5,154 metric tonnes of deuterium and a cruising speed of warp 3.0 and a supercruise of 4.1wf. Now, using the TNG:TM warp chart as a guide (because it's the only thing I can go on) those two speeds require 739 gigawatts and 18.7 terawatts, respectively. Plugging in those numbers and assuming a D+D reaction, that's 19.3 years and 9.3 months of flight power, respectively. In order to get the 0.6 months of flight he stipulates, you need so much inefficiency that the ship would melt from thermal loads.
There are other pure fusion designs (I don't remember which at the moment) that have the opposite problem: way too little power for the stated time if you assume D+D but go back to being melted slag if you assume catalyzed D+D.
However, I feel the need to reiterate that I have to cherry-pick examples that don't fit the curve. The preponderance of Rolfes' designs are quite good, technically. And, in general, I would suggest studying them as an example of excellent internal design work.
Sorry for the length of the post. I sorta warmed to the topic.
