I'm a fan of the OEM stuff. It continues to be developed and the latest is a Pi based system. It is expandable and hackable or just used out of the box. The original idea for a pi + enclosure, was to have an HDD attached to solve the problems with SD Card failure, but they seem to have solved the problem with a low write system so it is just SD card, plus a shield to do the clever stuff!

I'd encourage anyone who is interested in this to support them if you can.

https://www.kickstarter.com/projects/openenergymonitor/emonpi-open-hardware-raspberry-pi-based-energy-mon

Which is of course an argument against smart meters!

Of course it depends on what you want to do. This system can monitor temp, humidity, PV import. They have the backend to enable you to switch things via one wire relays and have an MQTT interface planned. I import my external weather data into it and ultimately I think I will be able to control my heating via it (if I ever get the time to play). As a basis for a whole house system I think it has much going for it.

Oh and the standalone version will run on batteries so you could stick it somewhere, let it run and gather data, then go and get that data.

It is also partly about how they try and run the company as well.

YMMV

i think you are missing the point of 'smart meters', they are destined to be used in conjunction with variable charging, and load-shedding during the day/night.

The ability to demonstrate to/educate users that the energy they use varies in cost during the day will smooth the way towards multiple charging bands throughout the day, combined with variations due to increased solar/wind inputs?

The nay-sayer on the radio just did not get the point, or was not willing to acknowledge it?

Initially smart metering was meant to be two way communication, but that is not going to happen any time soon.
If we, as individuals created stand alone systems that can mange the demand in out house, then there would only be a slight difference in the daily variance. Outside factors are what govern our lives. We currently have about 10 GW of nuclear power available, but our national baseload is very really under 20 GW. No amount of load shedding/time shifting is going to change this in the next 20 years. It is true that as wind and PV make a bigger contribution that there are some benefits from an environmental viewpoint in time shifting, but again they are limited.
As we move away from coal and replace it with RE and gas, we will have to load shift, and balancing will become more important and expensive. But this is 2 decades away.
Currently smart meters are only useful for reporting usage and seem to be 'locked' to one agency, the technology does not seem to be universal. British Gas jumped the gun on this one.
The ability to monitor in the home is useful to some people (I am one of them), but most people probably won't pay for the option to have a display. British Gas and EDF sent out millions of energy monitor, very few got used.

This is why I question the need for another monitor in the marketplace. if it is to succeed it has to be cheap and useful. So far, from my research (what my ResM is in after all) was that even knowing your usage, you are quite often helpless to do anything about it i.e. no good putting your lights on when the PV is generating or putting them on when you are asleep as energy is cheap.
We need to have a more integrated energy mix based on nuclear, wind and PV with gas taking up the differences. 3 of those technologies are very controversial and gas is susceptible to world energy price fluctuations. I don't know how much load shifting we can realistically do, probably no more than 15% of our daily variations, which yesterday (02/04/2015) would be 2.44 GW or about 100W per house (if only houses did all the balancing). It would be easier to save 2.4 kWh/day per house than try and manage it dynamically.
You could turn a medium sized town off for an hour a day, if we knew it was going to happen we would cope quite easily.
But the very easy way to sort this problem is to just put the price of energy up, demand reduction will happen naturally that way.

It seems to me that the emonPi is intended for people who already have some interest in energy monitoring for whatever reason and need to measure a few different parameters but don't want to mess around too much doing their own system. To compare it with either a home-brew Arduino (or whatever) system in one direction or a simple off-the-shelf energy monitor or smart meter in the other is entirely missing the point.