By features of size I mean per the ASME Y14.5M-1994 definition at 1.3.17. Per section 2.1.1.1 you should preferably use positional tolerance for locating features of size. That's my justification for suggesting it.
Your question on the overall surface profile equivalent to +-T brings to light one of the issues with lack of clarity on how the +-T could be interpreted when not applied to actual dimensions, but nominal geometry.
I was taking the +-T as a tolerance of how far a point can vary from its theoretically perfect location. As such I was proposing using an equivalent surface profile of 2T. As the machinist this works in your favor giving you the maximum tolerance zone. For example, on something like the length or width of the part, you actually end up with a tolerance of +-2T - although no point has moved more than T from it's 'perfect' location.
I think maybe you were applying +-T to 'features of size' and the like, or measurements from a pseudo 'datum'. In this case you apply a surface profile of T to give you an effective +- tolerance on the length or width of the part etc. of +-T.
It's your choice. If people aren't going to take the time to give me a fully dimensioned drawing, or explain how they expect their partial dimensioning applied, then I'd suggest applying the interpretation that gives me the most error budget so I'm more likely to make 'good' parts - so long as you make it clear in the contract this is what you're going to do.
On the position when you convert a square +- tolerance to position it's common practice to switch to the equivalent 'diameter' - at least for holes or circular boss's. This is based on the assumption that you have mating hole patterns with fasteners going through them or the like. For your position diameter you usually pick the circle that the square will fit into. The concept being that if the feature that sits in the corner of the square is acceptable, then it can be that distance from the center of the tol zone in any direction. So you end up with a circle fit on the corners of the +- square.
For example, say you have a zone of +-.005 (so a .010 square). Well Pythagoras theorem tells us the distance across corners is sqrt2 * the length of the sides. So in this case the 'equivalent' diameter is .014. Hence my suggestion to use a pos tol or sqrt2*2T.
thread1103-235569 talks about the topic of converting +- to position tolerance etc. in some detail.
It's easier to explain some of this graphically but I have too much work to create the required illustrations. Plus, whenever I do some pedantic wanna be checker usually picks it apart and that's not good for my blood pressure;-).
Let me reiterate, I think trying to have a +- overall tolerance without dimensions to apply it to is a bad idea as it's so ambiguous and as shown above can have multiple potential meanings.
Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
There is no reason to specify the same tolerance twice. The general tolerance is applied "unless otherwise specified", so the only time a specific FCF is applied is when it is needed for a particular feature or set of features that need more control that what is offered by the default. (I guess you can also use a local FCF is less control is required, but if that's the case, why bother with the specific FCF.)
Seriously though, the fabracator doesn't need to know something is important...they donly need to know what you tell them.
