From:  History of Yard Lumber Size Standards, U.S. Forest Service, pub. Sept. 1964

“Trend in Lumber Size Standards
There is strong competition not only among regional areas of lumber production but also between lumber and other construction materials. Survival of the lumber manufacturer demands the utmost efficiency. This economic pressure has been a compelling reason for the continuing erosion of standard sizes. Fifty years ago, 13/16 inch was a common thickness for the dressed 1-inch board, By 1929, 25/32 inch had become more common, the 3/4-inch board appeared in 1956, and the 5/8-inch board is now proposed. This latter value is no longer related to the nominal 1-inch thickness. Likewise, lumber dressed dry to 1-1/2 inches thickness does not require 2 inches rough green. The thinner boards and dimension are, of course, useful, and technical information to show their usefulness has been developed.”

The shrinkage in lumber sizes is partially, perhaps, to allow for drying and surfacing.  But it’s mostly just for economic reasons.  If the industry wanted to sell finished products at 2” x 4” (or if the government required it to), it certainly could.  And if you check the history as outlined in the above document, the industry used to at least come closer to true measures than it does now.  If, as observed above, lumber dressed dry to 1-1/2 inches does not require 2 inches rough green; then it’s fair to state that an actual board of 2 inches rough green will, dressed dry, yield a board thicker than 1-1/2.  If the industry is selling a 1-1/2″ board these days, it didn’t come from a “2″ board.  And it shouldn’t be labelled as such.  In my opinion, anyway.

A “2×4″ isn’t actually 2×4; but it is 1.5×3.5 and has been since 1961. Is it rather annoying that the product name isn’t its actual dimensions when it easily could be? Yes. However, the situation is stable and consistent, just not with the naive meaning of the jargon term.

Death after a few hours at as little as 0.16%, death within minutes at 1.28% (yes, those decimal points are in the right place).

If every single bubble in the bread was CO and you inhaled it you would die.

In this case, I think the expectations of how much bread expands/raises is too variable to be a fault.

From the Telegraph

Nine out of ten pint drinkers are being given short measures in pubs, according to trading standards watchdogs.

In several establishments, officers were shocked to find that almost 90per cent of the pints of beer ordered were short measure, with the worst pint having a shortfall of up to 11.8per cent — effectively a £0.40p overcharge on each pint of beer.

In a survey of 30 pubs, bars and restaurants, 88 pints were bought by Trading Standards Officers in Birmingham. Only 9 drinks were a full pint, making over 88per cent of pints pulled short measure. The average shortfall found was 3.94 per cent.

The prices for a round of 3 pints varied widely, from £6.40 to £10.55, with many rounds costing almost £10. However, the prices had little bearing on the quantity, with some of the most expensive having the greatest shortfall. In many cases, this equated to an overcharge of nearly £1 on a round of three drinks.

Neil Eustace, chairman of the City Council Public Protection Committee, said: “I am disappointed to see a rise in short measure beer and we are considering formal action in some of the worst cases. Customers should not hesitate to ask for a top-up if their pint looks a little low on festive cheer.”

The Weights and Measures Act 1985 states that a pint of beer should be exactly that and offences are committed for short deliveries.

(Also, I suspect that precisely controlling the length of baked goods not formed in mould-style pans is a bit tricky, so I’d be more concerned about being stiffed on the filling quantity, rather than precisely how far the roll slumped during cooking).

There is a WHO page for semicarbazide ( http://www.who.int/foodsafety/chem/sem/en/ ) , which mainly focuses on it leeching from plastics, and recommends that it be banned, especially from baby food glasses (which the EU followed up on some years ago).
Quoting from the linked page:

The presence of SEM has raised concern since it has weak carcinogenic activity when fed to laboratory animals at high doses.

Further studies aimed at clarifying whether SEM is formed from azodicarbonamide are needed. In addition, better information on the mechanism of toxicity and on exposure to SEM from all sources would help to define the nature of what is now considered to be a low risk to human health.

So, yeah. It might break down to something that might be dangerous in large amounts, and it’s easy to replace. Good enough reason to skip it, though it’s a long stretch to call it “poison”.

The only information I’ve seen on why is that the material itself is a respiratory sensitivizer – it can irritate the airways when inhaled, (triggering asthma attacks and such), which is reason good enough to ban it, especially when it doesn’t seem to be hard to replace or even manage without.

And if the combination of answers here seem odd: I really hate knee-jerk “omg chemicals!” scaremongering, and listing the alternative industrial uses for a product plus a list of (harmless) breakdown chemicals that sound a bit iffy to the uninformed is well into that territory. In this specific case, I agree that it shouldn’t be used – but given that the study in question was talking about factory workers inhaling the stuff, not people eating a bit of the breakdown products, I’d worry far less than what you’re aiming for.

Besides, you missed the main breakdown product, biurea. Its the one I’d be the most wary of, given that it’s an absorbed solid – though the evidence is that it’s rapidly expelled (through urine) and doesn’t do any harm.

It’s just bizarre, and more than a little off-putting.