Your film list has the following values attached to it from the "quality assessment" side of the database--which is more daunting due to its being less objective. (As these values have evolved, I've added multiple critical sources, including complementary weighting options for the "wisdom of the commons" data at IMDB--more useful for this purpose than what Wisconsin Mark seemed to think, at least for classic film--and several parallel averaging techniques to shield it from being too dependent upon my opinion):
711 Ocean Drive 86 e
99 River Street 90  e
The Accused* 81
Affair in Trinidad 77
Gilda 88  e
Lady from Shanghai 91  e
All the King's Men 89
Betrayed AKA When Strangers Marry (1944) 87  e
The Big Clock 85  e
Black Angel 84  e
Dark Corner 87  e
The Blue Dahlia* 81 e
Blue Gardenia 83
Champion 88  e
Brasher Doubloon 76
The Burglar 88
The Capture 83
Kiss of Death 87  e
The Chase 84 
Crime Wave 87  e
The Crooked Way 86  vg
Cry of the City 85 
*Films screening at NC 16
Films not shown at NC SF in italics
[n] rankings in revised Top 25 master list
"e" and "vg" from 2015 Solomon annotations made to master "Noir Top 25" list (full list)
That's a nice list, with films that are nothing less than enjoyable--and it's certainly got a higher average overall than NC 16--but there are really only two films that boil up toward the Top 25 there. That those two (99 RIVER STREET and LADY FROM SHANGHAI) aren't there is an artifact of the original poll that can't be band-aided sufficiently without three to four times as many participants in the revision exercise. There were just too many films that had been insufficiently circulated at the time.
The new Top 25 "ins" look as follows:
Big Combo 92 e, Act of Violence 92 e, Sweet Smell of Success 94 e, Vertigo 93 e, Pickup on South Street 89 vg
The Top 25 "outs" look like this:
Narrow Margin 90 e, The Big Heat 91 vg, Born to Kill 89 e, Murder My Sweet 87 e, Postman 84 vg
Your annotations again included here, to show how this group (interestingly enough) seems to form some type of dividing line. More on this at the bottom of the post...
Now the revision method seems to have admitted a film that is not quite as accomplished as a couple of the films that were previously in the Top 25 (PIckup on South Street). And that's going to happen once in awhile, usually because you don't have enough participants and/or enough other reliable data points to calculate a quality ranking.
Films that were missed/overlooked/not sufficiently seen in 2005 that now really look like they'd make a run on the Top 25 or 50 are as follows--just using the spillover from the NC list:
Une si jolie petite plage (94), Drunken Angel (94), Pepe le moko (93), The Breaking Point (93), Losey's M (92), Quai des orfevres (92), Stray Dog (92), The Reckless Moment (92), The Prowler (92), Diabolique (91), Rififi (91), Death of A Cyclist (91), Possessed (90)
And that doesn't even get into the foreign noir screened over the past four years at the Roxie:
Le jour se leve (95), Pale Flower (94), Voici le temps des assassins (93), Mollenard (93), Les jeux sont faits (93), Assault on the Pay Train (93), Seagulls Are Dying in the Harbor (93), Dédée D'Anvers (92), La verité sur Bébé Donge (92) Impasse des deux anges (92), Le septieme jure (92), Cairo Station (92), O Drakos (92), Voyage sans espoir (92), Retour de manivelle (91), Les eaux troubles (91), Chair de poule (91), Non coupable (90), Un revenant (90), L'assassinat du Pere Noel (90), Le monte-charge (90)...
This list confirms what ChiBob asserted about the state of affairs many of us here have experienced over the past four years as great "lost" films have surfaced, and certainly argues for the fact that a Top 100/200/500 would be much different (but much more authoritative) once all of the noirs from all over the world are able to be seen by a sufficient number of people. Indeed, a Top 25 list is truly a quaint notion--but we are simply trying to resurrect a portion of the evaluation process from what it can offer us, which is more than what we are getting from anywhere else.
Now, tying this in with your annotations to the original list: my count (which could be slightly off) suggests that out of the 186 films listed by the 2005 voters in the Top 25 poll, 102 of them grade out for you as excellent ("E"). That's about 55% of the list. That correlates to the most recent slice of the NC SF database in terms of its screening of films from the Top 25 poll: of the 63 films from the list that have played at NC, 43, or more than two-thirds, have a rating of 90 or higher.
Of course, that's just the "left side of the bell curve." And I think that's another way to state what you were pointing out. Now perhaps it is pointless to try to determine the additional difference in quality level once a group or list or class has been defined as "excellent." But it also might be interesting and ultimately valuable to do so, as we don't know what we might find out. That's why I suggested evaluating the ten films that switched places as a way to start. There was a consensus for a change in their value in the revision efforts; there is something that might a boundary line to be explored in those ten films, since your remarks suggest some gradations between your "excellent" and "very good" assessments and that they overlap in both film lists.
I will hold off from performing the exercise with the ten films that switched places on the two versions of the Top 25 list until some time has past, and others have decided to weigh in or not. That includes likely folks (ChiO and ChiBob, possibly Colin and Carl), those who don't ordinarily "rank" films in this way (the two Dans plus The Bitter One), occasional presences (JayM, Mike Kuhns, Ned, Gary Deane), and those who left via burnout or under some self-perceived notion of strained circumstances (Gord, Jefty, DT).
For now, I will just report on what my votes in the Noir Top 25 were for those ten films, reflecting what was everyone's exceptionally incomplete view of the noir landscape at that time:
The Set-Up 10 (the 16-20 range); Act of Violence 5 (the 21-25 range), Murder My Sweet 5; The Big Heat 1 ("honorable mention")
I wasn't very engaged in this "boundary"--not in the way someone like ChiBob was at that time:
The Set-Up 20 (6-10 range); The Big Combo 15 (11-15 range); Born to Kill 15; Sweet Smell of Success 1; The Big Heat 1; The Narrow Margin 1
Last note: the "left side of the bell curve" doesn't necessarily apply to a film-intensive program like NC (24+ films at a time; quality is, or at least should be, easier to control the fewer films you show). It actually has something close to a bell curve shape in terms of film quality; the curve skews left because there is a quality boost from the programmer's selection process. But it does still look like a bell curve...