Move Hash::Merge -> Hash::Merge::Augment, new Hash::Merge from Hash::Merge::F

Original Hash::Merge is incredibly slow due to the MONKEY. This is
someone that is very unlikely to be what people want. App::Assixt
reports parse time to go from 13s to 7s on first run, 13s to 0.3s on
subsequent runs, which is *massive*.

1 files changed, 1 insertions, 1 deletions

diff --git a/t/03-unit.t b/t/03-unit.t
index 43d70ff..5230398 100644
--- a/t/03-unit.t
+++ b/t/03-unit.t
@@ -2,7 +2,7 @@
 
 use v6.c;
 
-use Hash::Merge::Unit;
+use Hash::Merge;
 use Test;
 
 plan 2;