Bryan,
Thanks for drawing some of that out again.
I'm not sure why would you suggest I'm dodging the question of where the 117 acres were and what got swapped when I just put forward a VERY speculative suggestion regarding creating Golf House Road as a mirror of Turnbridge that frankly wouldn't support my theory very well?
I could have easily just said that you'd need to move the road about 30 feet along the length to the east, because I thought that is the dimension you had suggested earlier would yield a few acres in return from what P&H drew. If it's 60 feet, that's a possibility, as well.
As to your point about that making the "triangle" even less useful for golf, isn't that the point? It's only when you buy into the theory that the exact dimensions of the course were determined through some previous invisible, hypothetical routing that you run into that quandary and we already know that ALL of the rest of the course besides that western edge was bounded by historical boundaries...NOT anything purchased specifically for golf.
And, my theory is that once the 21.1 acre Dallas Estate was purchased in October 1910, that gave them a possible 119 acres of Johnson Farm they originally looked at in July 1910, + 21 acres Dallas Estate, + 3 acres of RR land next to the clubhouse M&W recommended for 143 total and we know they wanted to limit golf course acreage to around 120.
THAT is when the hypothetical road boundary got drawn by Connell and Lloyd, and the fact that it got tight up near the top because of the way the Haverford College rectangular property juts in is simply an accident of the configuration of the property; NOT because someone already planned some holes "up there". There was no "up there"...it was all just Johnson Farmland that had been further subdivided in an attempt to get down to the desired acreage.
I'm more than happy to try and answer your questions further, but I would ask you this first;
Is the 1910 Land Plan only incorrect and overstated in one measured area or is it uniformly and inconsistently inaccurate throughout?
In other words, does everything say, south of Ardmore Avenue measure out fine and everything above inaccurate, or is the entire map off by some percentage?
I would think that would be important for us to know, because otherwise, with the evidence I've seen to date I'd say that the answer is simply that the "Approximate Road" was misapproximated by some 20 yards as you suggest between what was proposed and what they eventually needed to fit the golf holes, and I know you're hoping I make a more daring, speculative answer than that!
Thanks.