Are you leaving the led on during sleep? If so, that's probably (part of) your problem. A typical led draws 20mA, which would drain the battery in under 28 hours. It seems the leds on the pinoccio draw a lot less, measurement show 1.2mA for green and blue and 2.4mA for red (I measured the resistors for each of the leds, which showed 0.6V and 1.2V, divided by 470 ohm).
At 1.2mA, the green led should be able to run around 19 days on a full battery. However, if you have anything else connected to the scout, or something else that leaks one or two mA, you're quickly down to just a few days.
In the past, we've found that the 16u2 being powered down can cause about 1mA of leakage current. This should be fixed by this workaround, but it's a bit dodgy. To check if this happens, you can measure the current over the TX resistor. This is the resistor between the "MOSI" and "SS" labels printed on the board. If this resistor shows any voltage (measured between both sides of the resistor), the workaround didn't apply and it is leaking current.
Additonally, you could try calling
Scout.makeUnsetDisconnected() after setting up pin modes in your setup function, to prevent leaking current on unused floating pins (this enables a pullup on all unused pins). IIRC this shouldn't affect sleep current, but can reduce running current by a few mA.
I also just noticed you're calling
SleepHandler directly? Why is that? This shouldn't be needed, just call
Scout.scheduleSleep(...) instead. AFAICS, calling
SleepHandler directly might be skipping the RF suspend calls, leaving the radio running during sleep.
Do you have anything to measure the current actually used?