When it comes to S.G readings there is no question that hydrometers are better than refractometers, as that is what they are designed to do. Having said that, I calibrate my digital refractometer to get the correction factor for every batch of new wort - yes the color makes a difference to the refraction index, not just the sugar content - and that is why a lot of people disparage refractometers as being inaccurate.
On the hot side, I use the corrected refractometer reading throughout the mash and boil to get my S.G., and take two hydrometer readings before and after the boil for verification. My converted refractometer readings are generally within 1 point. You do though have to be careful that there is no suspended matter in your refractometer sample. I sometimes use a tiny funnel with white filter paper. I still beats the effort to cool the wort for each hydrometer reading.
On the cold side, refractometers are the way to go, specially if you do small batches (5 gal) like me. I take a small sample from my conical every day (or twice a day) during fermentation, and at this point, all I'm looking for is trend, not accuracy. Once the Brix value stops dropping, you know you're done with attenuation. It also gives you a very good idea of hop creep during dry hopping, with very fine resolution without having to squint at a graduated glass rod in solution.
The only time I really value my lab-grade narrow-range hydrometers with temperature correction is when I'm checking finished beers for ABV, using it in tandem with my Refractometer. You have to make sure there is no CO2 in the sample.
The refractometer conversion calcs out there are numerous, and also built into BeerSmith, the software I use. The slight inaccuracy of digital refractometers vs. high-end hydrometers when trying to determine S.G. (even after fermentation, as long as you have a good I.G reading) is minuscule when it comes to the home brewing IMHO. We are after all going for repeatability, not absolutes. In a commercial environment, I can see how hydrometers make sense - the volume loss is inconsequential, and they are simple to use. Plus accurate ABV is critical for those guys.
Tilt - I'm on my second one (they swapped the first one out). I wanted to love it - great concept, and the software works great, including the tilt pi web reporting - but it has its problems. Besides krausen, dry hopping and any fermentation additions play havoc with it. (Maybe they need to think of enclosing it in a top-shielded flotation chamber attached to a conical temp-probe port, and it can run a temp controller instead of the TC probe). BTW, the temp reporting is spot-on. The Tilt is great for trend monitoring while one is away from the fermenter - I was recently on the other side of the world watching my gravity drop in my garage conical in real time. Unfortunately the bluetooth range through a steel conical is minimal. Your device - phone or raspberry pi - has to be pretty much up against the vessel wall.