FEAT: Performance improvement for &Array + scalar and scalar + &Array

* The new implementation avoids cloning the elements twice, and it
  avoids iterating over the elements twice. (The old implementation
  called `.to_owned()` followed by the arithmetic operation, while the
  new implementation clones the elements and performs the arithmetic
  operation in the same iteration.)

On my machine, this change improves the performance for both
contiguous and discontiguous arrays. (`scalar_add_1/2` go from ~530
ns/iter to ~380 ns/iter, and `scalar_add_strided_1/2` go from ~1540
ns/iter to ~1420 ns/iter.)

(Other changes to impl applicability removed from this commit.)

Co-authored-by: bluss <[email protected]>

Author: jturner314

src/impl_ops.rs

@@ -159,8 +159,8 @@ impl<'a, A, S, D, B> $trt<B> for &'a ArrayBase<S, D>
           B: ScalarOperand,
 {
     type Output = Array<A, D>;
-    fn $mth(self, x: B) -> Array<A, D> {
-        self.to_owned().$mth(x)
+    fn $mth(self, x: B) -> Self::Output {
+        self.map(move |elt| elt.clone() $operator x.clone())
     }
 }
     );
@@ -210,11 +210,11 @@ impl<'a, S, D> $trt<&'a ArrayBase<S, D>> for $scalar
           D: Dimension,
 {
     type Output = Array<$scalar, D>;
-    fn $mth(self, rhs: &ArrayBase<S, D>) -> Array<$scalar, D> {
+    fn $mth(self, rhs: &ArrayBase<S, D>) -> Self::Output {
         if_commutative!($commutative {
             rhs.$mth(self)
         } or {
-            self.$mth(rhs.to_owned())
+            rhs.map(move |elt| self.clone() $operator elt.clone())
         })
     }
 }