Pre-reduction folles of Heraclea - who can distinguish them?

[Maffeo]

I have a special interest in the folles issued by the Tetrarchy (from Diocletian till the last pre-reduction issue of late 307). Although I primarily focus on Aquileia issues, of recent I’ve been looking at those of Heraclea which are extremely collectable (at least in the sense of being both readily available and comparatively cheap even in the highest grades). Yet, my senile synapses start snapping in all directions when I look at the way in which specimens of such are classified, both on reference sites (such as the invaluable Wildwinds) and dealers’ sites (I’ve built up a large collection of pics of these with their alleged classifications). I get the impression that there’s very little in the way of clear thinking and that lots of sites just take a stab, guess at classifications instead of working them out properly.

The RIC list of pre-reduction Heraclea folles is this:
Group I: 294-298
12a Diocletian: A B Γ Δ Ε
12b Maximian: A B Γ Δ E
17a Diocletian: A B Γ Δ E
17b Maximian: A B Γ Δ E 
18a Constantius Caes: A B Γ Δ E
18b Galerius Caes: AB Γ Δ E
19a Diocletian: A B Γ Δ E S
19b Maximian: A B Γ Δ Ε Ѕ
20a Constantius Caes: A B Γ Δ Ε S
20b Galerius Caes: A B Γ Δ E S
23a Diocletian: A Γ
23b Maximian: Γ

Group II: 305-306
24a Constantius Avg: A B Γ Δ E S
24b Galerius Avg: A B Γ Δ E  S
25 Severus II Nobil Caes: A B Γ Δ E S
26a Severus II Nob Caes: A B Γ Δ E S
26b Maximinus II Caes: A B  Γ Δ E S

Group III: 306-307
30 Severus Avg: B Γ Δ
31 Constantine I Caes: A B Γ
First reduction...

Now, the first and last issues of Group I (12a, 12b, 23a, 23b) and all of Group II and Group III represent no difficulty whatsoever: they’re all distinguishable clearly and easily on the basis of mint marks and obverse legends. The problems arise with issues 17,18,19,20. RIC distinguishes these on the basis of ‘smaller heads’ (17 and 18) and ‘larger heads’ (19 and 20).  But when I look at purported specimens of such on the sites mentioned above there seems to be total confusion: for example what to me looks like a Constantius Caes large head (and hence 20a) often appears on these sites as Constantius Caes small head (18b) and vice versa - and so on and so on with the others. I’ll give an example: take the two alleged specimens of Constantius Caes smaller head, 18b, listed on Wildwinds – to me they both (especially the second) look like ‘larger head’ and therefore ought to be classified 20b. I could repeat examples of this endlessly from dealers’ sites where this confusion (even allowing for different dies) seems to reign supreme.
Can any kind person help with a ready way of distinguishing these issues?
Thanks.
Mike

[Vincent]

Mike, thank you for bringing this enigma to light. I, myself, often wondered about this distinction not only of this period, but  ones as well for Constantine and Licinius at other nearly mints of Cyzicus and Heaiclea. I even have one Camp gate of Licinius Jr. that looks like a small punch dot!
I'm not sure if this is related or not, but "The Celator" published an article that even examined the possible significance of the laurel wreath ties!
("Tales of the Ties: Laurel Wreath Ties and Their Significance in  late Roman Billion Coinage by Dr. Jogn Hardgrove and David S. Mitchell, issue 3, Vol 9 March 1995). The authors only concentrated mainly on Licinius Sr., as quoted, "Dr. Hardgrove counted a total of 15 different variations of laurel wreath ties on coins issued from Siscia mint alone. Furthermore, he found a similar variety of wreath tie designs at virtually every mint he examined.
...."Plainly, Sutherland and Carson considered the wreath ties to be an important feature, particularly in establishing the chronology of different issues."...."But the more Dr. Hardgrove studied wreath ties on coins in his collection, the more it became evident the variations didm't seem to follow a chronological order (!)". The author(s) make the possiblity that these were signitures" of the die engravers. However, perhaps a design of these laurel ties were employed on "small" heads and the other for "large" heads"? 
Anyway, worth looking at perhaps.