JW1SN family has 5VDC member as shown in his circuit.
i don't think this circuit will work on only 5V because of considerable
voltage drop on LED and relay is probaly not going to hold reliably.
Circuit is not worth simulating because relay is mechanical device
and factors like spring force, friction and holding current are not known.
if you insist on energy saving don't use relay and if you must, you will
also have to do some experimenting to get the values right.
diferent color LEDs have different voltage drop and if i recall correctly
lowest Vf is for red LEDs. this should work ok even with 5V but LED might need to go or it could be added in parallel branch as shown by audioguru.
to save power, series resistor for LED would have to be increased to
reduce LED current to lowest acceptable level (it affects brightness).
edit:
I'll take that back. Circuit and shown values do make sense but we won't know
for sure without actual test. I checked datasheet and it does show pickup and
drop out voltages (at least the max values). Coil current is some 100mA, and
this is why resistor was in parallel with LED. With coil resistance of 47 Ohm and
that current of 100mA, voltage drop would be 4.7V. LED will require some
1.7-2.2V to operate and can take 20mA (that's about 100 ohm).
With 120 Ohm in parallel and coil in series, whole circuit would draw some
50mA which could be enough to hold relay. Capacitor size is not critical
(doesn't affect current saving or circuit consumption) and can be increased
if required until relay can pickup reliably.