The two cathode lines on the schema were my invention merely to remind me about the double band.
Assuming Z100 is a 100V zener and Z150 is a 150V zener (which Google indicates they might be), from looking at various other switchmode schemas, I think the these two in series will be equivalent to a single 250V zener, and put back-to-back with the B3710 across the transformer primary, this combination probably makes a surge protector/transient killer. And while a bit charred by the blast from the melted IC pin, these all seem to test OK on the DMM, with voltage drop across each of them (at 9v, 25mA) 0.79V - same as the good ones elsewhere on the board. So I may just put these all back in and hope for the best.
The other two diodes are shorted. Google tells me that 24B2, which sits between Source and Control on the IC, may be equivalent to NTE5031A, a 24V zener. Another (much simpler) switchmode schema I found with the MUIP3E3 had a 10V zener in this position , although the data sheet for MIP3E3 gives the Control voltage (VC) as 8V, so I'm not sure why this zener would be 24V (or 10V). There isn't another of these on the board to make any measurements or comparisons unfortunately. Anyway, I guess no harm in trying this - it can't be more broken than it is now!
Which leaves the diode which, depending on which way up I hold it, reads "74" at right angles to ""90", or "06" at right angles to "47". There are some others of these on the board near 250V and 280V electrolytics. Voltage drop on these is 0.75V. Maybe another zener, or ultrafast recovery switching, presumably around 200V - any thoughts on this one?
Anyway, thanks for your help so far and any other insights you (or anyone else) have would be most welcome.