Personally, there are a few things I believe in that I've always found true.
1) most piggybacks themselves are not very useful because they simply can't control much. Because of that, you wind up chaining several together when you could buy a cheap standalone for the same price
2) Cheap standalones suck<PERIOD>. Either go all out on a great system, or don't do it at all. There are too many important compromises to go 50%.
To me, it comes down to what you can get the stock ECU to do.
If you buy an advanced piggyback that can control, for example, the air meter data, the ignition timing and/or coil<s> outright, extra injectors/lambda output/triggering devices/controlling extra devices - while tuning & logging information live, all at a reasonable cost, there is NO reason to buy any standalone!
Where people screw up is that they go out and buy some cheap, half assed $700-1000 stand alone. To get a top quality tune out of it you wind up with a little box that you either spend tens - hundreds of hours getting it to work right yourself, or paying hundreds/thousands to have done for you. Even then, it may, or may not have many capabilities you might want, and definitely won't have the wide conditions & safety features built into the stock computer. Buy a great standalone you're just left with paying for it in both time & money but not enough people actually do that.
What they wind up with, is spending all the time & money on it that someone smart with an advanced piggyback could get the same results with by spending a few hundred dollars, cutting the tuning time to a small fraction of a standalone, all while having many times the reliability & running conditions, AND still make the same amount of quality power.
Again, that assumes you can get what you need to happen, to happen with a piggyback. Most people SEVERELY underrate what advanced piggybacks can accomplish to begin with... If you can't get a reliable idle & part throttle running, or keep the A/F ratio in check by simply dropping in huge injectors & handling the air sensor, that's a problem for something like an AFC...
Something more advanced, you can do that, AND/OR add injectors to cover up any part throttle, or closed-loop fueling holes. OR modify the stock o2 sensor output to make the stock ECU cover up the fueling hole by itself!
There are also many tricks for most cars to get out of closed loop that no one takes advantage of. My favorite example for Toyota's is that if a car has an air meter, the TPS doesn't control anything tuning wise other than the ECU giving some light tip in ignition advance, or tip out retard... In reality, they do nothing more than trigger when to go from closed-loop, to open loop.
For example only, but I have used this on a few various OBD-I, and early OBD-II Toyota engine's.
Let's say the TPS sensor is a 0-5v scale, with 5v being 100% wide open throttle. Open loop mode is triggered at XXXX rpm, and/OR 80% of wide open throttle.
It's just as easy as wiring in a single throw, double pole relay into the TPS signal that breaks the TPS signal, and applies it's own voltage (Higher than the trigger point, say 80% of 5v is 4v, so you would obviously need something between say 4-6 volts. Not so high as to burn out the ECU tho!). Then you simply have the piggyback throw the relay by whatever means it can.
Be that RPM by throttle angle, or in our case, off a MAP sensor / boost switch. As soon as whatever trigger is hit, the relay throws and we're now in open loop mode running live instead off of oxygen sensor<s> that could lead to catastrophic lean-outs before the engine would normally reach open-loop mode. We also just threw the difficulty of part throttle tuning / ECU corrections out the window... There is nothing to tune out if the ECU is disregarding the o2 sensors to begin with!!!
What did it take do accomplish that? Maybe 10-30min depending on how slow you are at reading a wiring diagram to find the TPS sensor output, then splicing the new connections in. All at the cost of cheap wire & a relay.
On the downside, yes... There is another function modified by the TPS signal. That being the transmission logic. It will believe it needs to shift more aggressive for the same amount of throttle you give it when you've triggered open-loop than it would normally if you're under conditions that would be closed-loop. Ask yourself... Is that really a problem? No that's another BENIFIT!!!
This works on the few Toyota ECU's I've tried it on, but may, or may not depending on the application. One notable problem is drive-by-wire. You wind up in theory either confusing the ECU as to where the throttle actually is, or you inadvertently move where the ECU moves the throttle! Obviously in either case you're **** outta luck! Personally, I've never trie dit, but I don't see it working on DBW.
There is also no reason to need a separate unit to control timing / coil<s>. A good piggyback can:
intercept the cam & crank signals making the ECU advance/retard the timing intercept the dwell ignition signal to modify the ignition signal sent to the coil
simply control the coil<s> outright...
Advanced piggybacks should have modifiers for everything they try to do. Examples being make the base N/A air meter, and/or fueling maps run perfectly for a turbocharged engine while out of boost, but modify it by a MAP sensor run correctly under boost (or vice-versa, make a boost map, then detune for N/A).
Another would be having the same setup, but then scaling it to a temperature sensor. That way, you could either have it run rich while cold for emessions, or run rich while hot to save pistons. (Or both if you can completely program the algorithm)
In essence, a real advanced piggyback doesn't just quality as a piggyback. They also qualify as a cheap stand alone, as most advanced piggybacks can run an engine as a stand-alone.
I feel like I just wrote a dissertation, but people that don't have first hand experience using an advanced piggyback don't need to be telling people what they can & can't do. The possibilities are only limited by what problems actually arise, and how you choose to solve them.
Now personal opinion. I swear by Perfect Power's SMT's.
The kick the pants out of the same technical level e-manage, at 2/3 the price of a fully upgraded e-manage.
They kick the crap out of split second's stuff as they do much more, and tune/log live.
They also have multiple sets of maps. Want an economy/power setup? What about a power & more power setup?
No problem. Install a swithc to ground (Ooor if you're smoothe, take the ECT/Power button for the transmission on some models, have it also trigger a relay that grounds the SMT's map switch wire.) Push the ECT/Power button, and it swaps over.
I love my SMT-6. I setup one OBD-I (i.e. they don't throw/store codes for any reason because they're stupid) 3.0L v6 turbo (200cc stock) to run reliably with 680cc injectors by nothing more than tuning the air-flow meter. If I recall, it intercepted the small airflow-meter which we created a bypass for, ran the ignition timing . We set it up to run 100% N/A, but used a pressure switch to swap the unit over to it's second set of maps which contained boost friendly fuel. If I remember correctly it wasn't a boost pressure switch sensing positive pressure, we installed a vacuum switch taking manifold vacuum that would trigger at very low vacuum like 2-3 in/hg (i.e. high throttle). (We had to invert the signal with a relay)
That way, the SMT would be swapped over as the engine was about to finish spooling, instead of all-ready creating positive psi.
I had fitted 720cc's on it with a buddy, but we took them back off. Yes... They did work also.
AFA OBD-II, you don't get the "wide birth" freedom of tuning that older OBD-I cars have, but as long as you either KEEP the stock ECU reading acceptably normal ranges, or SIMPLY MAKE IT THINK that it is still close enough to normal ranges.
You can accomplish anything.
mfg: 
