Illustrate new failing scenario with multi-arg function identifiers

As far as I could tell, this behavior is only observed when the arity
  of the function is higher than 1. It's fine for single-arg functions
  somehow.

examples/acceptance_tests/108/lib/tests.ak

@@ -0,0 +1,44 @@
+use aiken/builtin
+
+pub fn reduce(xs: List<a>, zero: b, do: fn(a, b) -> b) -> b {
+  when xs is {
+    [] -> zero
+    [head, ..tail] -> do(head, reduce(tail, zero, do))
+  }
+}
+
+pub fn inspect_1(self: Data, result: result) -> result {
+  builtin.choose_data(
+    self,
+    fail,
+    fail,
+    reduce(builtin.un_list_data(self), result, fn(a, b) { inspect_1(a, b) }),
+    {
+      trace @"int"
+      result
+    },
+    fail,
+  )
+}
+
+test as_lambda() {
+  inspect_1([14, 42], True)
+}
+
+pub fn inspect_2(self: Data, result: result) -> result {
+  builtin.choose_data(
+    self,
+    fail,
+    fail,
+    reduce(builtin.un_list_data(self), result, inspect_2),
+    {
+      trace @"int"
+      result
+    },
+    fail,
+  )
+}
+
+test as_identifier() {
+  inspect_2([14, 42], True)
+}