So with your hints on where to look, I've been playing with the values and found some interesting stuff. You probably know all this, but as you may quite literally be the only person in the world who finds this interesting, I'm going to tell you about it.
I started out by playing with the phases in the turns.csv for the 2 person GC (I'm using your final naval mod btw, from 2017). Countries just join up too fast; it seems that the double diplomatic phases at the beginning plus the ease of massive swings causes Turkey (almost always), Bulgaria (quite often) and Italy (too often) to join in the August pre-turn. I don't mind that occasionally but not almost every game. The AI does a good job of going after the important minors, but it's just too easy to get them in the war early.
As there is no way to set a timer on entry (as in, they get to war level and then actually join some months later, as Turkey did in real life), I decided to try to force a later entry. I started by reducing the pre-turn to a single diplomatic phase; if you drop the events phase, the diplomacy (IMO) works beautifully, but you get no events from the warplans. And if you have only the events phase (no named diplomacy phase), you can't actually send ambassadors, despite having the button for it. Clearly, there is some custom code around that events phase that I couldn't see how to circumvent, so dropping a pre-turn phase was a no-go.
But then I found diplomaticlimits.csv; that is where the rules for the "cautious" diplomacy for certain minors is stored (max +-2 columns on diplomatic table in 1914). Using that as a guide, I just went even harder for that. I changed Turkey to have a max of +-3 columns in 1914, Italy to have a max +-1 column in 1914 and +-3 in 1915, etc.
Nation 1914 1915 1916 1917 1918
GER 0 0 0 0 0
AUS 0 0 0 0 0
ENG 5 5 5 5 5
FRA 0 0 0 0 0
ITA 1 3 5 5 5RUS 0 0 0 0 0
USA 5 5 5 5 5
TUR 3 5 5 5 5AFG 5 5 5 5 5
BEL 5 5 5 5 5
BUL 1 3 5 5 5CHI 5 5 5 5 5
DEN 5 5 5 5 5
GRE 1 2 3 4 5HOL 5 5 5 5 5
JAP 5 5 5 5 5
LUX 0 0 0 0 0
MEX 5 5 5 5 5
NOR 5 5 5 5 5
PER 5 5 5 5 5
POR 5 5 5 5 5
ROM 1 2 3 5 5SER 5 5 5 5 5
SPA 5 5 5 5 5
SWE 5 5 5 5 5
SWI 0 0 0 0 0
ALB 0 0 0 0 0
This was a good start, but just to really force what I wanted, I then removed the diplomatic phases from August and September 1914 in turns.csv. So you get the double diplomacy in the pre-turn, but you don't get another shot at diplomacy until October 1914. This seems logical to me; after the initial flurry, nobody was in a hurry to join until they saw how the war was shaping up (and were sufficiently bribed). Unrelated, but while I was editing turns.csv I also removed the Politics phase until Nov-Dev 1914; forming a new government and increasing National Will that early in the war makes no sense and may theoretically increase the chances of sudden early victory. Maybe.
Turn Year Start Month End Month Description TP_WarPlans TP_Redeployment TP_Events TP_Diplomacy TP_Reinforcements TP_Military TP_Politics TP_Interphase Western Front Initiative Eastern Front Initiative
1 1914 8 8 [IDS_PRE_TURN] 1 0 1 1 0 1
0 0 TI_CentralPowersInitiative TI_AlliesInitiative
2 1914 8 8 [IDS_AUGUST] 0 0 0 0 1 1
0 0 -0 -0
3 1914 9 9 [IDS_SEPTEMBER] 0 0 1
0 1 1
0 0 -0 -0
4 1914 10 10 [IDS_OCTOBER] 0 0 1 1 1 1
0 0 -0 -0
I'm just in initial testing so far, but it seems promising. In my first test game, after Initial diplomacy, I had FRA, RUS, ENG, BEL and SER vs GER and AUS. TUR is at level 7 and Pro-Central and will almost certainly join in late 1914, BUL, ROM and ITALY aren't that close to war at all in Aug 1914. Just about perfect IMO.
If necessary, I'll play with the starting levels and even the diplomatic table (removing the most extreme results and adding more 0 or small results). I may also try having fewer diplomatic phases for the rest of the war as well; if Romania and Greece are still joining in 1915 even with these values, reducing the bites at the apple might do the trick.