There is a lot of misunderstanding about servos and how they relate to gyros and other functions. This is being propagated freely on the net and at many flying sites. Generally this has to do with the differences between types of servos (analogue or digital) and the relative specifications of these servo.
First let me make it clear for this discussion, that I believe reliability of a servo to be the most important specification of any servo. Generally speaking, any servo of almost any reasonable spec, will not cause a helicopter to crash if it continues to function properly. Lower speced servos may increase one's required thumb movements when trying to fly more precisely, but only if it fails, will it be sure to cause a crash.
I would have more faith recommending a lower cost analogue servo, over a higher cost digital in any application, if it has been proven to be more reliable and is of adequate power. If the lower cost unit was slower and would get sloppy relatively quickly, but had a history of never failing, then I would prefer it to a high priced digital that is 3 times faster and has 3 times more torque, but has a proven high failure rate. If I knew that the high priced digital could not take the higher vibration level of a helicopter, for any reasonable length of time without failing, then I would choose the lower priced analogue unit.
But let us assume, for the rest of this discussion that we are talking of servos that are equally dependable, what are their differences? What are their advantages and disadvantages?
I will also assume that everyone understands that a digital of equal torque and speed to an analogue unit will be the better performer. The question is, can a digital of lesser spec be a better choice then an analogue of superior spec. The answer, in most cases, is a resounding YES. And it is not because the manufacturers are publishing misleading specs, it is simply because of how differently each type of servo operates.
A digital gets up to full speed and full power very quickly, usually within the first 1 to 3 degrees of servo wheel rotation. On the other hand an analogue servo will take from 10 to 15 degrees to do so. An easy way to demonstrate this to yourself, is to take any analogue servo that is installed in a heli, and turn the TX and RX on. Now try to move the servo. Even if the analogue servo is a top quality one, with 150 in-oz of torque, you will be able to move it a bit before it will resist you. Now try that on a digital, even a weaker one with only 30 in-oz of torque. You will find that you will not be able to move it at all. Right from the moment you try to force it, it will resist you with all of its available power!
To make it more simple to understand the torque and speed relationship, lets slow everything down and change the rotary motion to linear motion. As examples, let us use two servos, both servo's specs are related to 60 degrees of travel using a 4.8 volt power supply. The digital is rated at 30 in-oz / 0.22 sec and the analogue at 90 in-oz./ 0.11 sec. Initially you may think that the analogue servo is three times stronger and twice as fast.
So why then is the digital so much more expensive and why does it work so much better controlling the tail rotor with any gyro?
Maybe it would help understanding all of this using two cars, each with a single speed transmission as a parallel. This would be like saying the analogue car, after traveling 6000 feet would be traveling at 100 miles per hour and producing 300 horse power. The digital car would only be traveling at 50 miles per hour and developing 100 horse power.
But what is the relationship after only traveling 200 feet? Well the digital would already have accelerated to its top 50 miles per hour speed and be up to its' full 100 horse power. The analogue would only be traveling at 13 mph and developing only about 15 hp. The analogue car would take about 800 feet before accelerating to 50 miles per hour and 1500 feet before hitting its top specified speed of 100mph!
Obviously, if one is talking about using one of these cars in a race of only 1500 feet (e.g. ¼ mile) then the digital car would be the winner every time.
As most of our R/C heli applications do not require 130 in.-oz. of torque, but do require quick, powerful and accurate response from any point, a weaker and even slower digital will almost always be a better choice. It gives full power almost instantly and also accelerates to full speed almost instantly. These qualities are even more important when a servo is being used in conjunction with ANY gyro
When the gyro senses an uncommanded (undesired) yaw, it will be much more efficient if the servo it feeds the correction command too, does so immediately and with all the power it has available. If it has to wait for the servo to "ramp" up, it may be too late requiring another correction command and the cycle will repeat itself. Tail oscillation may result at even relatively low gain and/or delay settings. This will result in a less solid feel to the tail rotor control and the gyro will not be able to hold as accurately and/or with as much tenacity through hard maneuvers. This is why, a relatively inexpensive gyro, when used with a good digital servo, will be capable of performing better then some more expensive gyros that are mated to even high performance analogue servos. Why do you think gyro manufacturers, that also manufacture good digital servos, always make combos with one of their good digital servos?
Do not let yourself become a slave to numbers, they do not tell the whole story. There may well be more to performance then what meets the eye on paper.
All that being understood, let me say that for general sport flying and basic 3D, that a servo on the collective and the cyclic of most heli's (inclucing 60's) will be fine using a good dual BB servo such as the 9202 Futaba servos. The 30 to 46 size helis can even use lesser servos. Even the lowly S-148 Futaba servos have been known to fly the 30 size with adequate authority. But the harder one pushes their heli, and the more predictable they want it to fly, the better the servo will be required.
A servo that uses only one bearing and a bushing to support the output shaft, is not any better then one that has the output shaft supported by two bushings. This is because the single bearing will just cause its lower bushing mate to wear twice as fast, so slop can develop just as quickly as in the dual bushing unit, maybe even faster! A dual bearing unit is a must for any improvement in wear factor and precision. A coreless motor will be required for any centering accuracy and they do accelerate faster then do standard 3 or 5 pole motors. And the new Futaba BLS brushless motor series will have a much lower current drain then will any other digital servo.
Also note that for cyclic and collective, torque is more important then speed. For tail rotor, speed is more important then torque and acceleration is more important then both of these (thus the advantage of the digital in this application). The throttle does not require a very powerful servo, so even a slower one will work with adequate speed (little load). As resolution is not that important an issue on throttle, in a pinch, one can use a longer arm to get more speed.
Also of note is that a mechanically symmetrical set-up, along with symmetrical ATV settings will guarranty optimum performance from any of the mechanical or electrical control systems that we find in our R/C helicopters.
Considering the above, you should now be able to choose the best servo, within your budget, for your application