I would bet the Jalapeno SHOULD be more efficient as you don't also turn a fan. Any power not directly running the hashing chip is wasted in the calculation. Then again actively cooled components can use some less power if they are running close to a thermal limit.
That having been said No I do not feel there was an obvoius deception from BFL. I had hoped for an early ship date. I didn't order promptly though and depending on how big the first run is I may do better or the same because of this. I didn't intend to use my cash for the upgrade. I paid with bitcoins I made with my a little over 6 weeks to ship single I received in may. The only "deception" I see is people not actually reading the whole agreement. I find it hard to blame a business over people apparently rampant illiteracy.
Lets look at it a different way. Lets say I am interested in watching Psych. It comes on after Monday Night Football. Now the NFL says the game should end before Psych starts. Somewhere along the lines more time is spent on instant replays and time outs plus the game is a real nail biter and goes into overtime. By the time the game is over the show I intended to watch that the NFL promised me would be on as scheduled is half over. Was I decieved? Personally NO. Many years ago before I realized that Political speaches, Football, Basketball, Hockey and other spots do not end on time reliably and no padding was ever put in of sufficient time to cover multiple overtimes and time outs or a prattling on elected official. Did I hope to get to watch the whole episode sure, but I didn't expect to get to.
I look at BFL actually in a better light then my example. They stated it wasn't late until January 2013. That seems like a hint. It could be as late as the first of next year. Do I think they tried to actually get everything together. Yes. Have I heard so far reasonable explinations for why a more complicated chip might be delayed. Yes. Do I personally think that their fabrication house will eventually get a chip that runs as well as BFL expects. Yes. The reason I look at BFL in a better light isn't that their hopeful finishing time was wrong, but because they actually gave a hopeful end date and a worst case scenerio type end. Sports and elected officals never give me the maximum time it could get to.
Could BFL run over January. Anything is possible. I think it is unlikely that they would. Then again in 3 weeks I may change my mind based on new(ish) information and a lack of time for another attempt.
As far as the working chip you really do seem to have the cart behind the horse as is proper. Expecting a fabrication house to put out a working chip in 3 months doesn't sound that unrealistic. July, to August, to September to October. At least I think the preorders started in June but even then I am leaving a month out just in case I am wrong. So by October whatever day they actually would have ordered in July they have had 3 whole months to spin up a chip, ship it, test it, and order a run of those. Maybe the timeline is a little tight but I can't imagine a 3 month backlog on a chip design. This is assuming they hadn't started the order before pre-orders started. I can't say they did try to get a chip fabbed up in time for a test shortly after preorders, so that wafers could be printed sooner.
Right now no ones numbers are really hard numbers, I would say though that if the process nodes of known chips are correct no one seems to think the power is out of the realm of possibility. bASIC seems to be shooting for a hardcopy setup and a larger size. If bASIC has a working prototype why don't they have a hard number from their prototype? My guess would be that bASIC doesn't want to quote their current numbers. It could be that they didn't bother measuring or that they don't care to release the information. At either rate I would take BFL's guess over nothing from bASIC.
I think you're very forgiving of BFL's consistent do-over shipping dates, but that's your prerogative. I find it more than a little disturbing that they announced a product for pre-order so early in the year, began taking money for said product, and yet here 6 months later still do not have the primary component to build that product. Even when they do have working chips which I believe the latest estimate was mid Dec, there's still board layout, firmware, optimizing, testing, etc. Consider Tom's situation, if his releases are to be believed he has the chips, and he was still forced to announce a month and a half delay, but BFL is going to get the chips in mid Dec and ship by Jan 1? That doesn't even pass cursory sniff tests.
Don't really understand all the bASIC doesn't release power info talk anymore. I did think Tom was being a bit too quiet on the subject, but he's since said 80-120 I believe. A good size range to be sure, but if you don't have a final board, and you're talking about increasing chips and hashes then a wide range is prudent. Assuming the power consumption is 120w for 72Gh, I got zero problems with that, just get the dang thing in a box and ship it to me. A single day hashing at current difficulty w/ 72Gh would likely pay me the difference between the single SC's power consumption costs and the bASIC01's for the entire year. Not to mention the huge disparity in $/Gh between these two competing gizmos.