Some minor comments, but otherwise looks good

Reviewed 2 of 14 files at r1.
Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @pbardea, @RichardJCai, and @solongordon)

pkg/ccl/importccl/read_import_pgdump.go, line 543 at r1 (raw file):

		// the auto stats stuff will pick up the changes and run if needed.
	case *tree.AlterTableOwner:
		// ignore

Why'd this move?

pkg/sql/alter_table_owner.go, line 25 at r1 (raw file):

type alterTableOwnerNode struct {
	owner     security.SQLUsername
	tableDesc *tabledesc.Mutable

nit: s/tableDesc/desc

pkg/sql/alter_table_owner.go, line 40 at r1 (raw file):

	tn := n.Name.ToTableName()
	requiredTableKind := tree.ResolveAnyTableKind

This deserves a comment indicating that an ALTER TABLE OWNER applies to views and sequences as well, and that's why we're resolving any descriptor as opposed to a table desc explicitly.

pkg/sql/alter_table_owner.go, line 94 at r1 (raw file):

// checkCanAlterTableAndSetNewOwner handles privilege checking and setting new owner.
// Called in ALTER TABLE and REASSIGN OWNED BY.

nit: comment over 80 characters. Also, could you remove the "Called in ALTER ..." seems more effort to maintain than its usefulness.

pkg/sql/logictest/testdata/logic_test/alter_view_owner, line 44 at r1 (raw file):

# MATERIALIZED keyword must be present to change the owner of materialized views.
statement error pq: "mvx" is a materialized view

What if you say ALTER TABLE mvx OWNER TO ...? Does that work with Postgres? Either way, it's worth adding a test for that case

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @arulajmani, @pbardea, and @solongordon)

pkg/ccl/importccl/read_import_pgdump.go, line 543 at r1 (raw file):

I split this out from being a general AlterTable subcommand to its own statement (so it could also used for VIEWS/SEQUENCES) so it no longer a subcase of *tree.AlterTable.

Not sure if this case ignore is actually needed though.

pkg/sql/alter_table_owner.go, line 40 at r1 (raw file):

Done.

pkg/sql/logictest/testdata/logic_test/alter_view_owner, line 44 at r1 (raw file):

Good call! That indeed does work with Postgres, adding a case.

Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @pbardea, @RichardJCai, and @solongordon)

pkg/ccl/importccl/read_import_pgdump.go, line 543 at r1 (raw file):

Ah makes sense, I was just curious.