[MINOR][SQL][SS] Remove TODO comments as var in case class is discouraged but worth breaking it

### What changes were proposed in this pull request?

This patch removes TODO comments which are left to address changing case classes having vars to normal classes in spark-sql-kafka module - the pattern is actually discouraged, but still worth to break it, as we already use automatic toString implementation and we may be using more.

### Why are the changes needed?

Described above.

### Does this PR introduce any user-facing change?

No.

### How was this patch tested?

Existing UTs.

Closes #26992 from HeartSaVioR/SPARK-30337.

Authored-by: Jungtaek Lim (HeartSaVioR) <kabhwan.opensource@gmail.com>
Signed-off-by: HyukjinKwon <gurwls223@apache.org>

external/kafka-0-10-sql/src/main/scala/org/apache/spark/sql/kafka010/KafkaDataConsumer.scala

@@ -130,8 +130,6 @@ private[kafka010] class InternalKafkaConsumer(
   }
 }
 
-// TODO: consider changing this to normal class, as having mutable variables in
-//   case class sounds weird.
 /**
  * The internal object to store the fetched data from Kafka consumer and the next offset to poll.
  *
@@ -192,8 +190,6 @@ private[kafka010] case class FetchedData(
   def offsetAfterPoll: Long = _offsetAfterPoll
 }
 
-// TODO: consider changing this to normal class, as having mutable variables in
-//   case class sounds weird.
 /**
  * The internal object returned by the `fetchRecord` method. If `record` is empty, it means it is
  * invisible (either a transaction message, or an aborted message when the consumer's