NavList:

 Tony Oz you wrote:  Having the Newtonian+NH (the Natural Horizon) as the golden standard, where ANY movement however minuscule does not affect the apparent position of the horizon and the star, I want the similar behavior from an AH sextant.

I think is is most difficult to achieve with bubble AH because its' responsiveness massively depends on the bubble size. The bigger the bubble - the more responsive and accurate it is. But this big and accurate bubble is difficult to align with a star - no apparent center to watch for. The bubble small enough to be convenient with a star is far too sticky...

I'm following an attempt to create a bubble AH for a "classic" sextant, the experimenter had to attach a cell-phone's vibro motor to make small bubbles respond to the gravity better.

Tony

There must be an expression for what you’re trying to do.  Something along the lines of ‘Chasing the impossible’, or ‘Chasing the not worth bothering with’.  Bubble, pendulous, and gyro artificial horizons were developed to cope with the situation where the natural horizon wasn’t available.  In addition to the usual instrument error being greater than a marine sextant because of the greater complication, plus bubble stiction (negligible in my experience unless you have a really mucky bubble chamber or too viscous a liquid), you picked up various errors due to the nature of the thing.  Aiming was more difficult and therefore susceptible to personal error, if you go along with that concept (some people don’t).  Then if the platform was moving, there were acceleration errors caused by speed changes, heading changes, rhumb line steering, and Corioli’s acceleration.  Averaging helped to reduce the error considerably, but overall, you were lucky to achieve a greater accuracy than 3nm.  This was acceptable if all you wanted was to arrive close enough to your destination to allow a visual or radio navigation approach, but woe betide you if you descended towards high ground on the strength of an astro fix alone. 

Might it be possible to add a sextant star sighting app to a top-quality smart phone?  Probably, eventually, but not yet for a while.  You only need watch a group of people outside a pub pointing their smartphones in different directions all claiming to have identified the same star constellation to realise that.  With better sensors, it might be possible, but the chances are the app would use GNSS somewhere within its workings for timing, levelling, or long-term stability, so why bother?  Why not just use the GNSS position readout? That could be accurate to 10 metres. Unless it's just for fun. 

You also wrote: I wonder if a gyro AH is free from "unresponsiveness".

I've no idea, I would suggest a gyro AH would be very responsive to tilt, but unless a good one, they'd lie when subject to acceleration.  Ask one of our airline or test pilots, but I seem to remember being told that in the early 50s when fighter perfomance got ahead of instrument performance for a while, if fighters taking off in bad weather were going to crash, they invariably crashed to the left of the runway extended centre line.  The bank error was due to the gyros pendulousity and the pitch error was due to the pendulous vanes (which your gyro wouldn't have). DaveP