I think SQL Developer is predominantly about migrating the metadata, and sample data across, so that you have get an Oracle schema up and running quickly to do testing etc.

When it comes to unloading/loading large quantities of data (and comparing it) I would say you are best using tools native to the particular product.

For example, for volumes up in the 1TB range, I'd be doing some home-grown comparisons, for example, I could apply an MD5 to each column, and sum up the values, and then compare to the same query in Sybase. I've split the MD5 into 2 halves to keep the number sizes under control

with hashvals as 
  ( select 
      standard_hash(col1,'MD5') c1,
      standard_hash(col2,'MD5') c2,
      standard_hash(col3,'MD5') c3,
     ...
   from migrated_table
 )
select 
  sum(to_number(substr(c1,1,16),'xxxxxxxxxxxxxxxx')),
  sum(to_number(substr(c1,17,16),'xxxxxxxxxxxxxxxx')),
  sum(to_number(substr(c2,1,16),'xxxxxxxxxxxxxxxx')),
  sum(to_number(substr(c2,17,16),'xxxxxxxxxxxxxxxx')),
  sum(to_number(substr(c3,1,16),'xxxxxxxxxxxxxxxx')),
  sum(to_number(substr(c3,17,16),'xxxxxxxxxxxxxxxx')),
...
...
from hashvals

and then do the same against my original sysbase data.

This gives the whole power of SQL to you - I can do it across the whole table like above, or do GROUP BY to break it up into logical units etc