Which miner did you use?
You can try my cgminer fork, I got 2.7-2.8GH/s averaged per chip on my 20 chips 4-layers board.

Of course, you're not just paying for a piece of silicon wafer, but also for the months of design work, the cost of the mask set, as well as a risk premium.

Whats the heat like?

Ah .. interesting ... so "immediately" turns out to be about 10 days?
On a second thought - that's probably normal and okay. There was a long weekend (9/2) in between, so there are 4-5 business days between then and today. Unless you ordered a "overnight" or "second day" shipping it would take on average 3-5 business days. If they shipped it on 9/3 and you received it yesterday - that's okay and perfectly within the expected time frame.

As for October availability Dave had mentioned earlier: "we'll have plenty of reels available starting early October". That's why I didn't consider "availability" as a concern.

"Do they have chips" was never the question. Dave mentioned that they have enough in stock that if I wanted a reel I could have it tomorrow.

I'm getting concerned about that... I've designed a product around prices that were announced months ago and then rescinded, and I've been waiting approaching two weeks for $800 worth of chips that are "sorry they're expensive but they ship immediately" to be shipped.   My whole model depends on chips being available early in October and near the price I expect.  Bitfury chips are great, and I'm thrilled with them so far from a technical standpoint, but there's not enough margin in them if anything goes pear shaped :/

yes I would be interested ... in such a software, github?

One of the selling points was a "fork" of cgminer (I am a mod on IRC #cgminer and do testing of cgminer). Well established and always forked.

The software chainminer using "getwork" REALLY,  it seems even a year after it's release, you have a patch work solution for stratum, the major protocol of mining pools.

chainminer has no relation to cgminer. Putin's comments on he'll work on it when he can, shows his priorities.

I asked specifically if Dave was sending a board to Con and Kano and he said he would, he didn't, guess it's another fail or bad decision.

Actually mine was an interesting case. Chip 4 was hashing fine, but was not forwarding SPI . Thus  I  had to cut the traces of Chip 5 first and jump the SPI. Still no fish.
Wish I had a scope . I graduated from the univ, so missing all the nice bench equipment.

So I have potentially 2 hashing chips that could be donated.

I'm watching for updates on that too ... I hope they keep up that promise and do finally announce the pricing.
I really hope they won't ruin their so far pretty good reputation and take the Avalon/BFL route with everything happening "in two weeks"

Yeah, they do, but the Drillbit usbs (barntech GB) the usbs were $100 each and can be clocked as high as 2.5 or even 2.7 with proper cooling etc.  As far as I know its 2.7gh\s theoretical maximum per chip?

At least have the decency to say why you're asking this, mr. Integrity:

https://bitcointalk.org/index.php?topic=289316.0

https://hashfast.com/simon-at-work-one-week-before-tapeout/

"This is Simon hard at work on chip design during the run up to tape out. HashFast engineers occupied the first few cubicles in the Uniquify office, while the business team (Eduardo, John, and co.) set up shop in the back, by in the broom closet." 

link? I was under the impression that if a board has <16 working chips, it may be possible to overclock the others using the additional headroom provided by not running a chip (ie: 2/16 chips dont work -> 12.5% more power available to working chips. (am i correct on this, can anyone confirm?)