The solder blobs have a hole where the nichrome enters the blob. And it seems to be that the hot wire has caused the solder to recede.
This might sometimes lead to atty failure but in the under 2 months use bracket, most failures seem to be due to wire snap due to deposit buildup; I still think that this is the majority case.
As far as tin getting into the deposit, I dont't think this is because of the solder melting around the nichrome alone, but also some electro-chemical action - because only tin was found in the deposit - no copper, nickel or silver.
The tin component of the deposit (about 3%) is not likely to be burned off and may be part of the reason why the coil is not shiny. Whether this tin coating might affect the resistance of the coil over time is unknown, but quite possible.
Hence, I suspect that the coil resistance becomes reduced over time where the wire coils around the wick. So eventually the hottest parts of the nichrome wire will be the 'legs' (these will be the parts with the most resistance and now heated more by the greater current; particularly, half-way between the gunked coil and the solder blob. This is indeed where most failures occur, heat softened and possibly mechanically stretched also to breaking point.