Luckily for me, I work just beneath a specialist green bean importer / distributor.. I get all the fresh coffee I want (usually 1-2kg at a time) - and its usually roasted / ground the same day! Much better than ANY of the muck you can buy in a supermarket!
The thing with Google is that I know they are harvesting my data - and I am guilty of letting them do that. For that priveledge, they give me some great things for FREE - we have Google maps / streetview - including on my phone which runs a Google operating system (that I get FREE, remember when sat-nav was expensive?), A secure modern browser which is FREE, a fantastic search engine which is FREE, many of Googles services that rely on the harvested information are genuinely useful, etc etc...
In the case of LG in this story, the data was harvested WITHOUT permission or knowledge, overstepped the bounds of what might be acceptable and served no useful function for the end user (Most people would agree that additional advertising on your TV is excessive).
Google is very transparent about what they collect and how they use the info - LG has been dishonest and overstepped the mark.
Unless you know of some magical new OCR technology then OCR is NOT feasible for this type of job. For it to work with 4pt text the OCR software would be very inaccurate. Modern OCR software uses predictive technologies such as dictionary checking, grammar checking, near-neighbor analysis etc in order to get good results. It expects text within certain size constraints in certain fonts and of a certain quality. A SSL key printed at 4pt might get 30-40% accuracy at best. Then you would have to compare each and every character by hand - that means looking at two separate images to make sure the OCR is correct.
Much quicker to have it blown up and have a typist copy it by hand. A good typist could get 98% or above accuracy at a fair speed - and they would not need to look at two separate images.
Disclaimer: I work on the development of a document management system with OCR capabilities and have studied many OCR technologies as part of my work.
OCR is not a perfect technology. Especially on 4pt text with mixed characters and no "dictionary" words it can check against. It would be just as quick to have the data typed in as it would be to manually check it after OCR.
If you are using Linux then that is your issue. The Netflix workaround cannot use hardware acceleration thus heating your laptop up. You could use caffeine to automatically disable your screensaver whilst watching.
Once Netflix goes to HTML5 hopefully this will all be fixed!