Smyth

Weapons are no problem, they can be disabled by mission. Systems are the problem, and this is a general DCS issue.
If all modules had the ability to hard disable datalink, helmet sight, and GPS per-mission (and countermeasures for all the 70s aircraft in game with 80s-90s chaff/flare dispensers, and 90s TERNAV on the AJS37, etc. etc.), then it would be much less of a pain point for the community to be stuck with one specific example of each aircraft.

I understand that for a study sim game, making many similar versions of one airframe is economically unattractive to developers. However some small effort to mitigate this by disabling systems would go 90% of the way to enable an order of magnitude more immersive historical missions.

I had a similar issue and eventually noticed that the default keybind for both of those commands was 'LShift + D'.
At first the persistent red exclamation point would completely block the pinky switch radar selection on joystick buttons, and no combination of disabling/enabling/re-scanning devices fixed it reliably. After starting over and deleting those double-assigned keybinds first, the situation improved but was not completely resolved. Now I only have to go to control manager and choose 'rescan devices' every time I start the game, which has been working reliably. I don't know if that will work in your situation.
Edit - confirmed experimentally that this can be fixed completely by updating line 193 in "DCSWorld\Mods\aircraft\F-4E\Input\F-4E-Pilot\keyboard\default.lua"
From:
{ name = _('Wheel Brake (Left)'), combos = combo('D', 'LShift'), pressed = iCommandLeftWheelBrake, up = iCommandLeftWheelBrake, value_pressed = -1, value_up = 1, category = { categories.gears_brakes_hook, categories.pedals } }, To:
{ name = _('Wheel Brake (Left)'), combos = combo('D', 'LCtrl'), pressed = iCommandLeftWheelBrake, up = iCommandLeftWheelBrake, value_pressed = -1, value_up = 1, category = { categories.gears_brakes_hook, categories.pedals } }, Then fixing the .diff file in 'Saved Games' folder to work with that.
However this is still not a great solution because DCS updates will likely break the keybind again.
There must  be some reason that most players are not experiencing this, but I cannot figure out what it could be.

To learn more about this topic, one good place to look is USAF or USN flight manuals for older aircraft (F-5, F-4, etc). They usually include IRL performance charts at different altitude and/or load. The performance overall may be lower than newer jets like F/A-18 and F-16, but the principals are the same and the details are easier to find for obvious reasons.
Also I'll go ahead and try to summarize, because I've never written this down in one place before and it's interesting to try.
First we have to separate sustained/instant turn, Mach/IAS, and Weight/Drag because they do not always move together. Then note this is specific to a subsonic dogfight in a supersonic jet like the F/A-18:
Sustained turn (and acceleration/climb):
Altitude++ -> Turn rate-- , Climb rate--, Best Mach==, Best IAS-- Weight++ -> Turn rate--, Climb rate--, Best Mach==, Best IAS== Drag++ -> practically no impact (Negligible for light A-A missile loads; for heavy A-G loads the impact from drag might be measurable, but irrelevant compared to the huge impact of weight). Instantaneous turn:
Altitude++ -> Turn rate--, Best Mach++, Best IAS== Weight++ -> Turn rate--, Best Mach==, Best IAS== (best speed depends on assumptions about airframe strength) Drag++ -> absolutely no impact (not even on the amount of speed lost, because induced drag at ~30 degrees AoA is easily an order of magnitude larger than parasitic drag). Please note once again, some of this is specific to supersonic jets at subsonic speeds. Subsonic jets like an A-4 or F-86 will also lose performance with altitude and weight, but details like best speed move differently, and drag becomes important. At supersonic speeds, some performance measurements may go up with altitude, and drag is extremely important.
 

(Edited 9/1): Shortly after I posted this, @IvanK came through with some original EM diagrams for the F104G. Please see his post later in the thread.
 On the plus side, it turns out my sketchy linear fit got within 4% which is about as good as I could have hoped for. The conclusions remain the same, but I've updated the graphic below with the real data.
For a comparison to other cold war fighters, I've taken a graphic I was working on to compare official data for the F4E and Mig-23ML with existing 70s jets in DCS and replaced the F5E with the F104. Hopefully the format makes some sense.

The result is pretty much as expected, but it's nice to have a direct comparison instead of trying to eyeball charts in different formats, at different weight and altitude.
F-4E is just better at anything in the horizontal, but it's a Phantom after all (I will not shut up about F4 maneuverability). Against the Soviet fighters, an F-104 can technically match their sustained turn rate, but its extremely restrictive instant turn envelope would make that very difficult to use. Overall the F104G is the one opponent where a pre-MLD Flogger might want to take a subsonic fight, however the Starfighter still destroys it in subsonic climb rate and acceleration so even then maybe not.
The drastic drop-off in turn performance with the flaps retracted at M0.85 is also going to be a major issue. How much of an issue depends on how flap damage is modeled. There is a NASA document (TN D-6943) investigating F104 handling with the flaps deployed up to M0.94, and anecdotal evidence of an F104 pilot accidentally leaving takeoff flaps extended while supersonic without severe consequences. Ideally Aerges finds a compromise that encourages players to fly realistically, without imposing an arbitrary and non-physical limit.
Not exactly, if I understand your question correctly. That sustained turn is only for the flaps deployed below M0.85. Without flaps the sustained turn will be below the maximum G for a while longer, but I don't have any information available to extend that dotted line below M0.85.