Enable unchecked_tombstone_compaction on the CQL schemas

[Roguelazer]

We run read repairs on our cluster and were having trouble with old SSTables not getting dropped until we added 'unchecked_tombstone_compaction': 'true' to the compaction settings in the schema (which came from reading some threads about optimizing TWCS in the face of read repairs). What do you think about adding that to the default list?

Also, it looks like the tombstone settings from #257 still aren't in the schema?

[iksaif]

@uepoch / @Thib17 is this something you could look at ? thanks !

Otherwise I could validate a quick PR :)