Good to know.  I have 'unlimited' service (150 GB/month) for which I pay about $1400/year.  The cap is a summation of both upload and download I believe.  It drives me nuts when people have these adolescent fantasies about Bitcoin being robust 'because satellite'.  Nothing could be more absurd and in multiple ways.

The last time I actually helped the network was when I ran bitcoind on my i586 based 1/2G router.  When that became impractical due to traffic growth I never did bother to set up a port forward (and sure as hell was not going to run UPnP!)

If people could at least break even by providing service in the transfer network (which I mistakenly thought was on the drawing board when I read the whitepaper) I doubt that we'd be facing issues.  Addressing this issue would be the prime area I'd like to see focused on.  As I've mentioned, a dedicated sidechain to apportion reward in a 'segregated' manner seems to me the most likely way to achieve this.