Some documentation cleanup

BUGS

@@ -14,8 +14,8 @@ Although this should happen rarely enough not to affect performance, this can
 be an issue if your application can't afford restarting the transaction,
 e.g. because it interacted with external environment.
 
-Client always raise in tpc_finish if a failure happen
------------------------------------------------------
+Client always raises in tpc_finish if a failure happen
+------------------------------------------------------
 
 This is wrong because the failure may actually happen just after the transaction
 is actually committed. Client should not raise if it can reconnect and note

TODO

-RC = Release Critical (for next release)
-
   Documentation
     - Clarify node state signification, and consider renaming them in the code.
       Ideas:
@@ -16,9 +14,7 @@ RC = Release Critical (for next release)
     node where the most important changes are needed.
 
     General
-RC  - Review XXX in the code (CODE)
-RC  - Review TODO in the code (CODE)
-RC  - Review output of pylint (CODE)
+    - Review XXX/TODO code tags (CODE)
     - When all cells are OUT_OF_DATE in backup mode, the one with most data
       could become UP_TO_DATE with appropriate backup_tid, so that the cluster
       stays operational. (FEATURE)