My approach is one very much based on risk - how likely am I to receive requests from data subjects requesting deletion of their data? How likely am I to be subject to a targeted attack where people try to remove information from my server? How likely am I to be the subject to enforcement action if my server is hacked and data is leaked?

On one argument operating a blog is a purely personal activity and so out of scope of GDPR in any event. If you're outside the EU, GDPR will only apply if you are actually offering goods/services to those in the Union, or are monitoring them. I take the point about analytics in the second place, but in the absence of analytics, I don't see that making available a blog constitutes the offering of goods/services?