If Litecoin increases its price, perhaps the L3 is interesting 
The difficulty of the hash rate is increasing at a far faster pace. It is easily negating the value of the increase in the coin. The larger release of the units hasn't even hit the pools yet. LTC profitability is down almost 50% already from the stated highs. That will get much much worse. But it's a TON higher than it was up to the bit price runup, despite the recent hashrate/diff jumps.
I came VERY close to shutting down my A2 units late last winter because they were very marginal on profitability DESPITE my low electric cost, and I was contemplating adding "more efficient" rigs in their place.
Right now, they're bloody near my MOST profitable current rigs (though keep in mind I don't build riser-GPU rigs so all my GPU rigs are 3-card).
They also cost me ballpark HALF what most of my recent GPU rigs have cost me, and are long since paid off....
X11 total network hashrate IIRC is quite a bit lower as a "number of D3 units" basis than Scrypt is on a "number of L3+ or A4 units" basis - it'll probably be longer before Scrypt profitability drops to "last year" levels for an L3+ then it will be for a D3, especially with *3* folks making big X11 units now (and who knows what Baikal will come up with to replace their current chip with).
One kicker in the deck - BW.COM and their Scrypt miner, which is CLAIMED to be 28nm therefore quite a bit cheaper to design and make than 14/16nm that the L3+ and A4 seems to be based on, AND a lot easier to get chips for.
If BW.COM starts making and/or selling a serious number of THOSE miners, all bets are off.
Scary part - it's specs are better than the A4 and VERY CLOSE on efficiency AND hashrate to the L3+.
Perhaps THIS unit is why Bitmain cut L3+ pricing?
On the other hand, BW.COM had announced plans at one point to sell their Bitcoin miners, then ended up either not doing so or ONLY selling them in bulk to large farms.....
