I have been using spider type fiberglass poles for years. Best of coarse is when used as verticals and some insuled wire tie wrapped to the pole for a monoband vertical.
Your point about the bend moment when the coax is fed to the top for a dipole or inverted vee configuration is valid.
However the wire elements will act as guy supports and so far I have never busted one. Definetly useless if you plan on using them as the end supports, They won't break but you will be around 20 feet high with the resulting bend !
Also, I put a 4x4 into a post holed hole around 2 ft deep and tamp rocks and soil while plumbing the remaining 6ft of pole.
Some C type screw on pipe clamps are used to attach to the post.
Bob

Hello,http://www.eham.net/articles/28962
Might be useful.
Bob

Hope for that money your antenna plans exceed a dipole or G5RV, hi
Bob

the "fbi" virus is everywhere not a eham cause.
If first time seen ,safe mode and go back to earlier date and restart.
Norton will handle it, if installed after restart.
Bob

Give some thought to a fishing rod casting reel using a 2 ox weight, gets me up and over 60 ft trees usualy on first or second try.
bob

Don if you are thinking wide band 40 antenna with substantial gain and amazing front to back please see my qrz page.
Bob

Use the following link for cutting phase lines, the article you refer to was in my infancy of phasing, Bud developed this;http://www.va7st.ca/home.html/2009/11/christman-phasing-calculator/
It will help you a lot.
Bob

Yeah thats what I meant, hi
Bob

The best advice I can give you is to stick with 1/4 wl spacing using 84 degree phase lines and a 71 degree lag line for direction changing.
135 degree phasing is extremely tricky to achieve while matching phase and voltage in equal amounts at both antennas.
While in theory the 1/8 spacing and 135 has slightly better gain I stick with the quadrature, christman method and very happy with 24 db minimum front to back from 7160 to 7200 as verified daily with east and West dx stations along with stateside contacts.
Bob

A couple of electrical points, I will leave the durability issues to others more informed.

1; The edz will be highly directional at that height assuming little interaction with roofing materials, I mean that sucker is a bidirectional lazer so pick your aiming points accordingly. I had to build other 40 antennas to fill in the missing lobes, hi.

2; Cecil can give you an exact length of 450 ohm feedline that will give you 50 ohms at the rig, thinking 3/4 wl but I forget.

3; As an engineer you already know the trade off of weight vs duribility, in this case the sag, windload etc, I am concerned that the wire you plan on is overkill but alas I don't live below you, hi.

Good luck.
Bob

OR; your 20 meter dipole is your most efficent antenna, noise is a signal also!
Bob

Welcome, check out an article here on eham http://www.eham.net/articles/29061
Might be helpful.
Bob

Rob that is one of the best explanations of verticals I have read anywhere!
Good solid information.
Bob

Reality check, there is no way 30 ft of wire or alum. in any configuration is gonna be a multiband efficent antenna, no matter what extensive ground system or tuner you might add.
Amazing how the advertisments try to forget the laws of physics.
Most times your feedline is your emmiter when off anything beyond 40 meters, think Tak antenna and the isotron, it is what it is, and good for limited expectations when small area for real antennas.
Bob

The original question asked if a 25 ft dipole, rotatable would be better for dx than, insert by me; A Well Designed Vertical with substantial near field ground loss reductions and nice far field reinforcement.
Answer still is, and will be the vertical.
That low of a dipole is not directional except for non dx signals or might be a recipiant of a nice 4 element mono bander up 100 ft over in Italy doing all the heavy lifting, not my definition of dx.
Get the dipole up beyond a 1/4 wl high and then it might exceed the vertical on most days.
Bob