vulnerability to compromise?

[DEMOREP1 at aol.com]

Mr. Ossipoff wrote-

Compared to other non-Condorcet methods, Approval
requires twice as many mistaken compromisers in order to give away
an election. So one disadvantage about compromise strategy is the
possibility of giving away an election. Additionally, it sends the
wrong message to the other voters when you vote Compromise > Desired.
No one will know how many people consider Desired their favorite.
---
D- Since Condorcet 1, 2, etc. rankings exist, then Approval should be deemed 
at most only an interim measure to help get rid of plurality in the single 
winner case (despite Approval's * possibility of giving away an election* to 
a true lower ranked choice due to miscalculations of which choice gets the 
earliest majority if true rankings were being used- a true first choice 
majority, a true first plus second choices majority, a true first plus second 
plus third choices majority etc.).

The standard stuff-

Majority > Minority
Divided Majority > Divided Minority
Part of Divided Majority plus all or part of Divided Minority = Compromise

Since rankings are used in any accurate/ semi-accurate multi-member 
proportional representation method, why can't rankings be used immediately 
for the single winner case (which is the limiting case of the general 
multi-choice case -- such as electing N of Z choices) ???