It's an interesting idea and one that I might be tempted to support, but I'd have two concerns:
1. The write-up (at least as far as I've seen) comes across totally as an Arduino project and not a weather station project, ie maybe I've missed it, but I can't immediately see anything about sensor specifications, suggested (radiation) shields, and recommendations about adaptations for long-term robust operation in cold wet climates.
2. I'd be really quite surprised if a 0.5W panel charging a 1Ah battery could genuinely sustain reliable 24/7 operation with eg continuous 5 minute updates throughout winter at higher latitudes. The Davis Connect, for example, is AFAICS a pretty similar concept in terms of architecture and power requirement (ie a microcontroller-based ultra-lower power design for the data acquisition board and a GPRS modem that spends most of its time asleep) and requires a 5W panel charging a 12Ah battery (both of which may need supplementing in more 'tricky' locations).
But maybe someone can put me right on these two aspects.