About the quasi-identical thing...
I don't know if it is the structure of the sentences themselves that make a quasi-identical's work difficult to read, rather it is the topics and opinions they choose to discuss.
For example, lets use INTj and INTp.
The INTp will write about predictions, hypotheticals, foresight, etc. with an emphasis on facts, knowledge, methods, practicality, etc.
The INTj is not about intuitive foresight, but wants logical connections and the various ways and alternatives that can be used to maximize this knowlege.
When reading INTp stuff, the creative for the INTj is activated to provide alternatives that can be used in every one of the INTp's predictions. (INTj thinks: "This guy just doesn't see the big picture")
The facts, knowlege, and practical methods of the INTp are grating on the INTj's psyche, for the INTj simply needs to know how all these things follow from one another, and the benifits that each piece of information provides. (INTj gets frustrated from having to use Ti creatively for each and every fact.)
In otherwords, the INTp will use facts to prove his predictions of bad events, but the INTj will use alternatives that show that these bad events are actually surmountable, explaining it with logical proofs.
Well, shouldn't an ENTj's stuff be just as hard to read for an INTj?
Maybe, but the facts for an ENTj have more importance to the things that they may write about. Instead of having to counter 100's of different randomly applied facts with his Ti, the INTj may only need to show the possible inaplicability of a select key few of them. (That is, an ENTj's argument is more likely to be based on one or two key facts which are quickly connected by the INTj.)
This is my take on the whole quasi-identical thing...