You can't remove chips, but I would imagine you can remove boards.
Your earlier complaint was rediculous, if you didn't realize it. You are essentially complaining about free hashing power.
Now, if you can't handle the added electricity, then fair enough. But I'm pretty sure these boards (not chips) just plug in.

 

Your belive that you can expand the capacity by factor of 3 from already pumped-up case design is ridiculous. My concerns were about amount of heat generated not about juggling with chip boards. The only reasonable explanation is that they don't have the case yet but that cast a big shadow on July delivery. EU regulations concerning the PSUs is another issue.

Does it really matter?  No company taking payment in full up front before a working product has been demonstrated has ever delivered what they promised, when they promised it.  This company will not be the first.  And why would they?  As long as people continue to be willing to pay cash up front for a product that has 0 chance of being delivered on time (seriously, how many times have we seen this EXACT pattern ), there really is no reason for them to change their business practices. 

See though this is exactly why (well, one of the many reasons anyway) no company taking preorders in this fashion will ever deliver as promised.  All those things you mentioned can of course be done.  But they also cost money, money which would then come out of their profits.  Since their customers have already paid and have almost no real recourse, why would they do that when they can just skimp on everything and when it doesn't work, push back the delivery date again and again?  We as customers have already demonstrated time and again that no matter how badly these companies fuck us, we're still happy to wait in line to give them our money for a product that will almost certainly never turn a profit.   

My friend, you have just won the internet

Also heard on Coin Desk  http://www.coindesk.com/private-china-meeting-bitcoin-mining-industry-leaders/

"The battle for share of the scrypt-coin networks is about to heat up too,
with at least four companies (Zeus, Gridseed, Innosilicon and Silverfish) apparently all ready to roll with existing inventory.
At stake is about 200 gigahashes-worth of total litecoin/dogecoin/etc mining power around the world,
all still running on GPUs – but not for long."

Will the last person mining scrypt coins please turn out the lights.

I vote no more pre order i think its time you companies really start taking on the risk now that crypto currencies are proven they are not going to disappear.

Pre-order is a great startup tool but the mainstream customer is taking on to much of the risk here and i don't think its really acceptable practice these days and lets face it,
you cant hide behind the excuse anymore that you cant get funding!

1-5% deposit refundable if you do not meet projected delivery dates and or hashing, power specs and lets face it, if a company starts doing that then so will you!

I  wouldn't compare the first ASIC's to the current ones regarding efficiency. They have learned a lot from those that have come before them and made efficient steps along the way.
If you constantly look to the past for your answers to the future, you will end up back where you started. Bitcoin, if anything, is the antithesis of that.

I'm curious as there seems to be a lot of hate going on here.

How about posters post in their threads whether or not they even have an order? I have a feeling next to none of you (except for those that have already openly stated it) have orders with A.T.
So the larger question is, why is what AT does so interesting for you all? Are you playing the role of Robin Hood, or rather are you representing other ASIC companies or

After all, its about sharing...

The suggestion works if you look at Cointerra and HashFast who both hopped on the newer ASIC technology bandwagon and promised astounding results. They started running into similar issues as AT with timeline, last minute changes, etc. Of particular note, Hashfast also flew by the seat of their pants on the chips at the last minute (like AT is doing), and it was their downfall.

Last I checked into it, Cointerra came to within about 85 or 90% of their original design spec, but still ended up late. Hashfast died off, kept taking money, kept making promises, failed to keep them.. and they're facing multiple lawsuits right now. Back when it was in the planning phase, I wagered that Cointerra would come out on top and Hashfast wouldn't. Honestly, Hashfast's CEO looked like a wannabee nerd. My wager paid off.

Here's some info on Hashfast's legal issues: http://arstechnica.com/tech-policy/2014/05/meet-hashfast-another-bitcoin-miner-manufacturer-accused-of-fraud/

I truly hope AT examines my suggestion from the previous page, and takes it seriously. There's 2 red flags in play right now against them making a July shipping date and there's enough undercurrent that they need to handle the situation.