IEEE 1584-2002 states that the formulas of Section 5.6 should be used for calculating arc-flash incident energies where current limiting fuses Class L and Class RK1 are used and the bolted three phase short-circuit current is within the appropriate current ranges.

The formulas of Section 5.6 are based on tests at 600V, 60Hz and a working distance of 455mm.

Question 1.
Do the formulas of Section 5.6 apply for 440V, 50Hz ?

Question 2.
What is the manufacturer of the fuses used to develop the formulas in Section 5.6 ?

Question 3.
The formulas of Section 5.6 are based on tests at 600V, 60Hz and a ‘working distance’ of 455mm.
The corresponding ‘gap’ for a ‘working distance’ of 455mm is 25mm for Low-voltage MCCs and panel boards.
Do the formulas of Section 5.6 apply for :
- ‘Gap’ of 25mm ?
- Arcing fault in a ‘box’ ?
- Earthed ( grounded ) system ?

Question 4.
The bolted phase fault current ( Ibf ) in the formulas of Section 5.6 are symmetrical rms current.
What effect does the actual system X/R and therefore current asymmetry have assuming the worst-case where the arcing-fault occurs at a voltage zero. ie. max asymmetry of arcing current ?

Question 5.
The formulas for incident energy in IEEE 1584-2002 Section 5.6 do not include arcing time.
How is the arc-flash boundary calculated when the arcing time is not known ?

It sounds like you are attempting to model current limiting fuses. The one thing I've learned about fuses over the years is that the modelling is either trivially easy or hard (iterative) and not necessarily highly accurate, depending on conditions. For the questions you are asking, I suggest contacting Mersen. I don't know if they have the answers either but knowing that several of their research scientists have been contributing information to the development of understanding of arc flash and have been providing key input to the IEEE 1584 committees, they would be the best source.

Do the formulas of Section 5.6 apply for 440V, 50Hz ?
No. Instead you'd have to model the fuse and then use the general empirical equation to determine the incident energy. This is actually how the various software packages do it.

What is the manufacturer of the fuses used to develop the formulas in Section 5.6 ?
I think it's Mersen but not 100% sure. Regardless they should be able to tell you exactly who it is if it wasn't them.

The formulas of Section 5.6 are based on tests at 600V, 60Hz and a ‘working distance’ of 455mm.
The corresponding ‘gap’ for a ‘working distance’ of 455mm is 25mm for Low-voltage MCCs and panel boards.
Do the formulas of Section 5.6 apply for :
- ‘Gap’ of 25mm ?
- Arcing fault in a ‘box’ ?
- Earthed ( grounded ) system ?
Not 100% sure. The goal of those formulas was to make estimates for current limiting fuses based on a set of fixed conditions. I don't know what the conditions are but as I just stated if you simply model the circuit using traditional methods instead of trying to use the empirical formulas in this section the above variables can be adjusted.

The bolted phase fault current ( Ibf ) in the formulas of Section 5.6 are symmetrical rms current.
What effect does the actual system X/R and therefore current asymmetry have assuming the worst-case where the arcing-fault occurs at a voltage zero. ie. max asymmetry of arcing current ?
X/R has a dramatic effect on arc stability (purely resistive circuits tend to extinguish). No way to tell without the actual test data, if it exists. If the data isn't on the CD that comes with the IEEE 1584 standard, you'd have to contact the source that contributed the data at this level of detail. I don't know what test conditions they used.

The formulas for incident energy in IEEE 1584-2002 Section 5.6 do not include arcing time.
How is the arc-flash boundary calculated when the arcing time is not known ?
The formulas in section 5.6 are meant for determining arc flash, not the boundary. They are meant for use with current limiting fuses, where simply knowing bolted fault current is not enough. Even then, IEEE 1584 is known to be not exactly the best way to do things with fuses. Several attempts have been made and currently the most likely method that the next version of IEEE 1584 will use is the "time domain method" which is published (minus a bunch of key details) directly on Mersen's web site.

1. I see no problems applying the formulae to 50Hz system. Please consider IEEE 1584 empirical model suitable for frequencies of 50 Hz and 60 Hz when in doubt. Note also that the IEEE 1584 fuse equations were developed based upon testing at 600V. I would not take liberty applying them at different system voltage including 440V.

2. The IEEE 1584 states "These formulae were developed [...] using one manufacturer’s fuses". Looking on list of IEEE 1584 sponsors, it is quite obvious the formulae were derived using Bussmann product. However, there are more than one Bussmann current-limiting Class L and Class RK1 fuses and the guide does not mention what particular fuse families were used for testing. Therefore, I would not take chances applying the formulae even to Bussmann fuses. The IEEE 1584 statement "Fuses from one manufacturer were used, but results with other manufacturers’ fuses of the same class should be similar. The manufacturer should be consulted" indicates the authors are not quite confident that the fuse equations apply to Class L and Class RK1 fuses. I don't personally find the IEEE 1584 fuse equations helpful at all. In my opinion, the manufacturer had taken advantage of being a sponsor marketing its fuse products in IEEE 1584 guide.

3. IEEE 1584 indicates the formulae were developed for an arc in 20x20x20 in. size switchgear. The guide also indicates 1-1/4 in. gap for LV switchgear.

4. Low voltage fuses are usually tested at 15% power factor associated with prospective peak current of 2.3 x the rms fault current. You can confirm it by looking on fuse peak let-through current characteristics. The 2.3 factor was chosen as a practical limit for low voltage circuits. Keep in mind the equipment X/R may have major impact on calculated available fault current if the ratio is ignored when calculating available short circuit currents. Please check this forum thread at http://www.arcflashforum.com/viewtopic.php?f=8&t=3391 for more information.

5. The arcing time is kinda built-in into the equations by means of fuse operating time as function of arcing current.

I would consider using IEEE 1584 Empirical model instead of IEEE 1584 Fuse Equations for arc flash analysis in systems protected by current limiting Class L and Class RK1 fuses.