I think the Schliemann could definitely be a lot higher. There's potential in there, especially considering just the sheer amount of prep required as White to stay on top of everything. Even the d3 lines (the "safe and positional" lines) have insane stuff like this:
The highest rating level you can get away with various openings against.

This is silly. The Italian, Marshall, and Sveshnikov are seen all the time at the 2700 and 2800 level (Indeed, with the Marshall, the only reason it's not seen more is that *white* is avoiding it). And any number of those others make occasional appearances. Just last month, Wei Yi won a masterpiece in a classical game at Wijk aan zee in a Vienna (via a Bishop's Opening Move Order).
The KIA tops out at 1800? Really?!?

Good luck playing the Grob against a 1850. If they are a true 1850 they should know how to take advantage of it.
Where is the London System? 😀
Good point.
I have entered the London System at 2300, this puts it at the same level as the King's Gambit and slightly lower than the Sicilian Dragon and Philidor. This may sound high for a system that a lot of average players take on just to avoid opening theory, however just because of players trying to avoid opening theory with it doesn't mean it's bad, even Carlsen has used it seriously. The main problem with the London System is its drawish nature. But I think if you learned it really well it might be good all the way up to 2300.
I noticed that the Sokolsky opening was not addressed as well.
I'm not too familiar with the Sokolsky. I had a brief read about it and put it at 2000, with the assumption that there is a tangible point to it, like the Nimzowitsch, but not fundamentally flawed like the Grob. If GMs have and sometimes still play it seriously it can't be unplayable at 2000. I also changed the Grob down to 1750 in response to the comment about it.
QGA 2700, Scandanavian 2450.

I noticed that the Sokolsky opening was not addressed as well.
I'm not too familiar with the Sokolsky. I had a brief read about it and put it at 2000, with the assumption that there is a tangible point to it, like the Nimzowitsch, but not fundamentally flawed like the Grob. If GMs have and sometimes still play it seriously it can't be unplayable at 2000. I also changed the Grob down to 1750 in response to the comment about it.
that sounds fair. I have played it against 2200 rated opponents but it can become quite drawish. I am still curious how you would access the other two openings that I mentioned?

Where is the London System? 😀
Good point.
I have entered the London System at 2300, this puts it at the same level as the King's Gambit and slightly lower than the Sicilian Dragon and Philidor. This may sound high for a system that a lot of average players take on just to avoid opening theory, however just because of players trying to avoid opening theory with it doesn't mean it's bad, even Carlsen has used it seriously. The main problem with the London System is its drawish nature. But I think if you learned it really well it might be good all the way up to 2300.
Gata Kamsky (2500+ GM) makes regular use of the London iirc.
@1
This is nonsense. Most of what you listed has been succesfully played in classical time control games by top grandmasters.
Bongcloud is a joke and is only for bullet troll games.
Parham attack (2. Qh5) - 1000
Hikaru played it three times in classical games vs GMs IIRC, scoring 2 draws and a loss (again, IIRC).
So you're a little off.
Also the engine eval is close to equal. So for that reason too you're not putting much thought into this.
-
Damiano's defence - 1100
I remember Fischer only scored a draw against this in a simul once.
I already addressed how using it as a total surprise is not what I'm talking about.
I believe Hikaru was trolling or having fun even in classical, a lot of fame these days is caused by being outlandish and getting in the chess news, becoming a chess personality is a lot more lucrative than actually being that good at chess these days.
I suppose you mean a computer evaluation, which often means very little in human terms. A human evaluation of equal or "close to zero" is disastrous for a white opening. GMs win about twice as much playing as white as they do as black.
I thought it might be interesting to discuss the highest ratings you can get away with using each opening against.
By "get away with" I mean that you can use it profitably as one of your main weapons and on a continuous basis, and the opponent knows you may use it, and you aren't getting punished for using it. The level it would hurt if you played it every game, even playing it well. Carlsen throwing in an occasional Italian or Alapin for surprise value is extremely different to using it as a main opening.
By "rating" I mean FIDE rating. I know different time limits also make a difference.
Here's what I've come up with, just from my interpretation, I am totally aware some of them are probably wrong, a little bit guessing sometimes. Feel free to disagree or add your own.
Parham attack (2. Qh5) - 1000
Damiano's defence - 1100
St. George's defence - 1500
Marshall Defence - 1600
Ponziani - 1700
Grob - 1750
King's Indian Attack - 1800
Latvian - 1900
Blackmar-Diemer Gambit - 1950
Goring Gambit - 2000
Nimzowitsch (1. b3) - 2000
Sokolsky - 2000
Owen's defence - 2050
Smith-Morra - 2100
Bishop's Opening - 2100
Vienna - 2150
Evan's Gambit - 2200
Bird's Opening - 2250
King's Gambit - 2300
London System - 2300
Sicilian Dragon - 2350
Philidor - 2350
Accelerated dragon/variant - 2400
Dutch - 2450
Schliemann - 2450
Scandanavian - 2450
Pirc - 2500
Scheveningen (5. ...e6) - 2550
Sveshnikov - 2600
Italian - 2600
Marshall Attack - 2600
Classical Sicilian - 2650
French - 2650
Caro Kann - 2700
QGA - 2700
King's Indian Defence - 2750
Najdorf - 2800
Bongcloud - 4000+