Aggregate (count) function forces query to use primary index only (even where there is a 2i index on the field in the group by clause

cbq> select count, capped_small from `bucket-1` group by capped_small;
{
"requestID": "e6071bb0-de55-48a0-9ff2-d13b83a6205c",
"errors": [

{ "code": 4000, "msg": "No primary index on keyspace bucket-1. Use CREATE PRIMARY INDEX to create one." }

],
"status": "fatal",
"metrics":

{ "elapsedTime": "1.720223ms", "executionTime": "1.335946ms", "resultCount": 0, "resultSize": 0, "errorCount": 1 }

}
cbq> explain select count, capped_small from `bucket-1` group by capped_small;
{
"requestID": "69078f8b-0127-4623-9772-6cdb0ec84f7a",
"errors": [

{ "code": 4000, "msg": "No primary index on keyspace bucket-1. Use CREATE PRIMARY INDEX to create one." }

],
"status": "fatal",
"metrics":

{ "elapsedTime": "1.936496ms", "executionTime": "1.577497ms", "resultCount": 0, "resultSize": 0, "errorCount": 1 }

}
cbq> select * from system:indexes;
{
"requestID": "517b4cec-3909-45c1-b968-005f0ed65532",
"signature":

{ "*": "*" }

,
"results": [
{
"indexes":

{ "datastore_id": "http://127.0.0.1:8091", "id": "4c757db390196977", "index_key": [ "`capped_small`" ], "keyspace_id": "bucket-1", "name": "by_capped_small", "namespace_id": "default", "state": "online", "using": "gsi" }

}
],
"status": "success",
"metrics":

{ "elapsedTime": "257.405389ms", "executionTime": "257.017194ms", "resultCount": 1, "resultSize": 431 }

}

any user would be confused with the error message stating a requirement for a primary index when they use an aggregate function, INCLUDING, explain on a statement including the aggregate function.