Hmmm. I agree with the point about it being tough. TOS Enterprise has its points. In TOS the Phaser was considered to be something of the 'ULTIMATE ' weapon....which was drastically down graded in capability over time. From being able to destroy half a continent to what??? Being effectively only a water pistol. This is why the Enterprise had only two Phasers - it didn't need any more than that...( subcontext: the Klingons weren't too familiar with Phasers at the start ( sub-subcontext: 'The Star Trek Spaceflight Chronology ' has something very interesting in it, an implication that proper defense screens, were quickly developed, thus leading to vastly increased Phaser power [Star Trek Phase II]
The TMP version of the Enterprise is of course a beauty. But really? Eighteen Phasers??? See above.
TNG is a proper long ranging ship. Designed to stay out for a maximum time of one hundred years, never to return. Planting Colonies along the flight path.
The Enterprise-C(Yesterday's Enterprise) is obviously designed to a more industrial esthetic. But seventy years limited. Just as fast as the Enterprise-D. But smaller, also dropping Colonies off. Under, ultimately control of Section 31?? Why? Star Trek Generations. They knew. And without telling the crews/dependents about the Borg. Probert's Enterprise a fantastic beauty, essentially the same as Yesterday's Enterprise.
The Enterprise-B... pity that they redesigned the Excelsior... the Excelsior class was is a perfect follow up to the Refit Enterprise... and had the range and speed ( warp factor twelve maximum safe cruising speed) to really penetrate into far distant spaces, especially looking for the Borg. In one year, going what would have taken the TOS Enterprise eight years to achieve.
The NX-01 Enterprise. As I see it, as stated the first really long range ship. Never mind the Akira class, math solutions are math solutions. The most likely mission profile that should have been, was to go straight out for one year at warp factor five, to investigate the first most unquie Stellar object. Statistically speaking. Then, go at an approximate right angle and do the same. Then return. Total mission time three years. This makes the Intrepid class, a science survey type of ship. They would have been launched, all of them in 2147 A. D., a total of twenty of them, to recalibrate the sensor systems, and to verify, and brench-mark the hand scanner. You don't send the NX class out empty handed... rhis mission profile of the Intrepid class, with twenty members, would be able to survey all stars in a twenty light-sphere, in one years time frame, assuming warp factor 3.6, allowing for a travel time of one month and a one month survey of the next system. It would have been very ordered in nature, but not to Vulcan standards. Why? Because it was based upon statistical logic, not machine logic.
Statictical logic involves the probability of something interesting, where 'something ' has either been seen very few times, or never before. Machine logic, has no particular interest in any volume of space - no uniqueness. All of the the same interest. Humans use filters. Vulcans don't.
In other words I love them all.