How do you know how accurate your Yaesu is? You are comparing an unknown to an unknown. Just because it is new and an amateur rig does not mean the counter is anywhere accurate enough to determine 20 Hz. I have seen brand new straight out of the box rigs be off over 100Hz.
It's similar in comparison to a simple, dead - Alarm clock - the alarm clock is right 2X a day even though it doesn't move, or even have its spring wound to keep time.
Compare the non-working Alarm clock to a watch that loses 1 second a day - The Alarm clock has more accurate "results" twice a day - because by the time the watch catches back up to itself - will take longer amount of time to be "corrected" it will be off for the rest of your lifetime.
T = 12 Hrs (figuring Analog watch dial)
T = 12 X (1 Hour equals 3600 Seconds) 3600
T = 43,200 Seconds (Hint: The Error)
Convert to Years = 365 (skipping Leap) Days = 1 Yr
Watch loses 1 second per day = 365 seconds per year
D(Lifetime) = 43,200 Seconds / 365 S-Yr
By the time (SIC) watch can even be correct = (118.3561643835616 Years) / 2 = No consideration for AM or PM
OR = D = 59.18(Rounded up) Years to the second of that moment
Compare to the Yaesu and the CB (Galaxy?) Now both radios may be "off" to the rest of the world but are in tune with each other - making it more accurate than most other radios what were never "matched" to these two.
Did you see my error?
Look more closely...there are 2 - 12-hour cycles for a dial watch or any face clock - being a day is 24 hours long.
So T = 43,200 X 2 or T= 86,400 / 365 = 236.72 years it will take...
So even with the demonstration - the error is even greater as others focus more on the accuracy of the measurement - which exacerbates the problem and multiplies the division of known actual time-base events versus the relativistic events - with relativistic events being two radios in-step with each other while time-based results of the argument being not every radio can be "tuned" the same way nor can it be considered accurate due to the environment and conditions it is operated in.
In light of the discussion - the effort to make all timing systems the same - generates the error within the time base used in each individual system - forcing the user to abandon the multi-point timing references to a one single time base as the reference