Since we know the current table name, it is redundant to name the table's primary key with the its table name.
Redundant but useful - I am a huge proponent of USING clauses in joins:
pk_tbl JOIN fk_tbl USING (pk_tbl_id)
As for "finding the PK" - its almost always the first column in the table, ends in "id", and has a reasonable prefix.
Being able to "grep pk_tbl_id" is also nice in many cases. Grepping "id" provides no useful value - though if one always uses table prefixes then grepping "pk_tbl.id" would mitigate that particular problem.
