Consider removing alias count limit

Per @korikuzma was initially in place because of some tricky DrugBank data, but now that we're using a different distribution of the data, it might not be an issue. Maybe either remove the limit entirely or just make it accord with whatever ends up being the 400kb size limit set by Dynamo.

No file, test, or entry point is identified. First locate where the alias count limit is enforced and determine how the alternate data distribution relates to Dynamo's 400 KB size limit. Done requires a decided limit strategy and evidence that the resulting alias data remains within the storage constraint.

Written by the indexing model from the issue text.