The downside is, controlling the activity is nothing like tuning a guitar, but more like pin the tail on the donkey.  Trends between structure and activity can be rationalized somewhat after the data is collected, but is wildly unpredictable beforehand.For the vast, vast majority of active pharmaceuticals, there is a reason the attrition rate in clinical trials is so goddamn high; toxicity is nearly impossible to predict.  Small changes in the molecule will cause different metabolic pathways, new compounds generated in vivo, which may be active with entirely unpredictable activity.Related to this issue:  there’s a famous example where kids were designing a street drug, and ended up giving each other Parkinson’s.
http://en.wikipedia.org/wiki/MPTP 

http://www.bbc.co.uk/news/uk-scotland-17735988 

Cause if so (and I honestly have no idea), I think it’d see a lot more, and more radical, use than you might be envisioning. You have no idea what I’d do to get a significantly more-perfect stimulant/nootropic effect in a single compound I could produce at home. Also, more finely-tuned mood-balancers.

Perhaps with the introduction of carefully-controlled software ‘apps’, similar to the ones available from Apple, we could see consumers have access to a personal drug designer they could use at home to create the medication they need.

is a bit disingenuous (I rather suspect Cronin’s closing comment was elicited to give an otherwise unremarkable press release some media spin, which appears to have worked).  Let’s face it, just the commercial, regulatory and legal issues associated with basically providing a largely uneducated public with the easy means for mass production of pharmaceutical-grade compounds would sink any such venture from the outset.

It could also be really nifty for producing drugs that aren’t commercially viable.  It’s not in a big drug company’s best interest to produce a perfect product for treating something a teeny tiny fraction of the population has.  So even if they run across something that might be promising for treating that condition, it’s not going to get a ton of research funding, and it’s less likely to get actually manufactured and bottled and sold and shipped.  The small printing concept might allow for making those drugs more viable.  Smaller labs and smaller producers could make small money making them or selling the formula to people or small producers to print them on their own. 

Automated reaction systems already exist, and are definitely not going to be generating pharmaceutical reagents in peoples homes anytime soon, for a multitude of reasons: number of reagents necessary, complicated conditions, purification!

Why would you want to store the 1000′s of reagents necessary to generate 100 pharmaceutical agents when you could just store the desired compounds in the first place?  Customized, personalized combinatorial generation of active pharmaceutical agents is not how medicine, or drug discovery works.  Research, yes, but the success rate of a molecule becoming a drug is far, far, far below 1%.

I can’t even imagine a reality where this is an efficient, practical solution, short of the Star Trek replicator.  So perhaps it is just one step closer to the replicator, but the only new or novel aspect of this is marketing towards the home user versus laboratories.

Or maybe I’m missing something.  I guess there’s nothing wrong with fictional extrapolations of current technology in order to drum up interest and excitement in science.

It’s not going to work as a “DIY medicine lab”. It’d be a complete waste to even try to use it to knock up a bit of asprin when you’ve got a hangover.
The printer ‘inks’ will need to include all the reagents. For medicines, this would be a bit more than CMYK pigments. The number of reagents would vastly outnumber the number of medicines that could be produced. 

Also, this is only going to work for some specialised reactions. There’s a reason most chemical plants have thick-walled pressurised vessels, vacuum pumps, steam lines and heating systems. You can’t make everything under ambient conditions and I wouldn’t trust a 3D printed reactor to survive very high/low pressures.

You also need some sophisticated analysis equipment to make sure the reaction worked and that any dangerous impurities have been removed before you even think about swallowing the product! Not to mention all the hazardous (flammable, corrosive, toxic, sensitising, carcinogenic) waste that needs to be disposed of safely (and the far greater waste produced when trying to make portion-sized syntheses).

Much of practical chemistry is analogous to cookery, so treat this as you would treat the announcement of 3D printed cookware. It’d be great for Heston Blumental’s novel cuisine but your standard cooker, pots & pans are going to be a lot more useful for everyday cooking.

Seriously, you think this thing would ever fly? The *AA’s screamed bloody murder when home taping devices were introduced. Could you imagine the reaction of a truly powerful industry? Geeze, it’s almost impossible to buy an eBook that isn’t crippled by DRM. In at least one state you can’t even buy chemistry-set glassware without registering with the government.

Between the neo-Prudes and the “rights holders,” this technology would never be allowed for, you know, people.

Also, yay for all the recreational drugs we’ll be able to invent.

Perhaps with the introduction of carefully-controlled software ‘apps’, similar to the ones available from Apple, we could see consumers have access to a ersonal drug designer they could use at home to create the medication they need.

Heh, carefully-controlled software that’ll be hacked and patched within hours ; )

Will this all occur on an L-5 space city? Will we have to wash our homemade miracle drugs down with liquefied soylent green? Will it be 115 degrees in January the day we first get these machines in our “kitchens”?