I will not depart on a flight if the destination is fogged in, until the fog has actually cleared at least to approach minima, regardless of what the TAF says about when it should clear.
Even on one of your famous 7 hours flights? The fog will certainly have dissipated by the time you arrive! 
The case being discussed must have been one of the lucky 10% http://www.aviationweather.gov/adds/metars/?station_ids=eddh+eddp+eddb+edve+edbc+edah+koak&std_trans=standard&chk_metars=on&hoursStr=past+18+hours&chk_tafs=on&submitmet=Submit
At least in the US, the forecasters are discouraged from writing CYA TAFs (the “unamendable TAF”). For the commercial operators, an overly pessimistic TAF is a nuisance (can’t dispatch), just as an overly optimistic TAF is (need to divert). So the TAF is usually not the doomsday scenario but the forecaster’s best judgement. It would be fun to dig up some statistics comparing TAFs to METARs; I bet the FAA and NWS actually keep a pretty close eye on those.
dissipation of a persistent condition (fog) is optimistic at least 90% of the time.
Not in my experience. It’s wrong most of the time by at least a few hours, but both ways.
The fog will certainly have dissipated by the time you arrive!
Not necessarily, I’ve seen fog that didn’t dissipate for almost a week.
Another gotcha with TAFs is that worsening cloud base need not be reported unless it crosses a (rather low) threshold. For example, a ceiling dropping from BKN090 dropping to BKN020 need not be reported, which is ok for IFR but can get you into trouble if you expect an easy VFR flight. See this explanation (in german).
Since the TAF was also issued by the national weather service, that “2nd opinion” won’t contain extra information
This I do not agree with. There are several reasons why.
In the TAF, you have a certain amount of information you can put inside and quite straightforward rules how to implement them. Also, they are usually not done by a briefer, but by a forecaster, at least here. So whoever you have on the phone if you call in, he will be able to give you a lot of additional information which by the very nature of it can not be put in the TAF.
The briefing staff have the whole lot of information at their disposal which the forecaster uses to generate his message. There are nuances which can be present but do not fulfill the trend criteria (trends are only published if they do fulfil some criteria, as a consequence, quite a lot of information is lost in translation so to speak.
Just as an example, I remember vividly a discussion I had here about the fact that a TAF gave CAVOK or NCD (can’t remember which one) while there was a solid overcast. Trouble was, the overcast was over the MSA, therefore for the purpose of the TAF, there were no clouds given. Other things are how to interpret the trends, how “sure” things like a Tempo 40 vs Tempo 30 is, what the TEMPO means (often, single events, what else may happen which fell trough the rules), how they expect the BECMG to develop (Vis 2000 m BECMGxx0812 5000m) and so on.
Taf Interpretation rules are different yet again According to most rulebooks I’ve seen, legal TAF interpretation always assumes the worst, ignores certain keywords e.t.c. Some samples:
BECMG161218 If the trend indicates an improvement, you can only expect it after 18z. If it’s indicating worse than prevailing conditions, you have to assume it will at 1201z. That is why I don’t like long periods such as this one of 6 hours and that is why they are not done very often.
TEMPO: In the end, you have to take TEMPO as constantly. Yes, TEMPO TS means the TS comes and goes, but for planning, you need to PLAN with TS.
PROB— TEMPO is usually disregarded in flight planning.
and so on, one can fill books about this.
By all means do get additional info from the Met office, it will give you additional slants on what they really think, not only what they put in the TAF.
An airport near me recently reported this METAR:
ESCM 170650Z 00000KT 6000 2000N R03/0700VP1500U R21/0750VP1500U BR BKN003 05/05 Q1027=
I understood this to mean that RVR was (at least from time to time) lower than the lowest meteorological visibility! I didn’t see how this could be possible, so I called the MET briefer to check that my interpretation was correct and apparently it was.
It seems that even though “2000N” should mean that the lowest visibility is 2000 m (and to the north), apparently the viz can in places be even lower. In this case there is some fog drifting around, but then I would have expected BCFG to be reported.
I guess I have to dig into Annex 3 and read the detailed rules. Or the MET observer has simply made a mistake.
RVR is Variable between 700 m and more than 1500m, and the trend is Upward (on RWY 03)
The prevailing visibility is 6000 m, except to the north where it is 2000 m. The prevailing visibility may not be observed at runway level, so if you have ground fog it may be quite a bit higher than RVR.
Isn’t visibility measured at the same level as RVR? (3 m above ground?)
In any case, I feel that the METAR should have included BCFG or possibly MIFG.
Prevailing visibility can be observed from the tower which is significantly higher.
This scenario happens a lot. Most of the time, the met station is not on the level of the runway, as you need an overview of the airport, so it has to be 3-5 m at least above airport level. So yes,if you have a 2-3 m thick fog layer over the runway then you can exactly get this effect. The observer can see 6 km, the RVR sensors see considerably less.
I´ve even seen situations where only the runway and the immediate vicinity was in fog, whereas the tarmac, tower, everything else was perfectly in the clear. Or even worse, where slant view from the airplane showed the runway perfectly until 100 ft or so, when it disappears totally. That is why RVR measurements and if possible camera systems are so important for observing.
I agree however that it is very weird that the present weather was not BCFG or PRFG. MIFG, possibly. BR does not do this situation justice.
As someone who does this all the time, these kind of situations FG/BCFG/PRFG/MIFG as well as low stratus are amongst those which really make us work hard, as well as Thunderstorms/Showers. Most of the rest is workable.
Most of the rest is workable
in switzerland, southwesterly flows seem to be notoriously hard to predict, too.
