primary key(fk_author_id , blog_id),

would be

primary key(fk_blogger_id , blog_id),

create table blog(
fk_blogger_id int(10),
blog_id int(10) auto_increment,
….other columns,
primary key(fk_author_id , blog_id),
unique key (blog_id),
foreign key (fk_blogger_id) references blogger(blogger_id)
) ENGINE=InnoDB;

“blogger” is a separate table of bloggers.

Now, say i have say 1-2 blog inserts every second posted by a random bloggers.
would it be a efficient table design?
I would prefer this design as it groups blogs per blogger. Direct access of a blog will use the unique index of blog_id and that would result in two index lookups. But as i would show other blogs of same blogger on the page, The grouping would speed up the overall efficiency.

Am I thinking right? are there any benchmarks around similar solution

great stuff though!

As Xaprb said, your second insert is slow because you are inserting in a random order. Before loading, sort your data in primary key order and the two inserts should run in similar time.