Fix NODENAME env var in calico kube-proxy.yml (#338)

In the calico kube-proxy.yml file, the daemonset was setting the
NODE_NAME env var to spec.nodeName, but in the kube-proxy start.ps1
script, it was looking for a var called NODENAME:
https://github.com/kubernetes-sigs/sig-windows-tools/blob/3ed28295f8694dffd03e144af4423137b2b8c30b/hostprocess/calico/kube-proxy/start.ps1#L61

If the env var is empty (as was always the case with this difference),
it defaults to the hostname, which might work in a great ammount of
cases, or thus is my suspicion for it going undetected.

Using the same env var name in both setting it in the yaml and reading
it in the ps1 script fixes it.

hostprocess/calico/kube-proxy/kube-proxy.yml

@@ -27,7 +27,7 @@ spec:
         name: kube-proxy
         imagePullPolicy: Always
         env:
-        - name: NODE_NAME
+        - name: NODENAME
           valueFrom:
             fieldRef:
               apiVersion: v1