Among total error of 3.1%, Cepheid reddening is 1.4% and the second largest source of uncertainty. SN Ia statistics is the largest with 1.9%. Rarely discussed in popular treatment is anchor distance, the third largest source with 1.3%. It is uncertainty of bottom lungs of the ladder, eg the distance to Large Magellanic Cloud before Cepheid and SN Ia are involved.
Yes, and I'm sad that it's so rare.
The Megamaser Cosmology Project is incredibly awesome; combining line of sight acceleration, velocity, velocity gradient, and observer angle on the sky is very "anchor".
https://safe.nrao.edu/wiki/bin/view/Main/MegamaserCosmologyP...
Gaia seems to be making good progress on direct parallax up to kiloparsecs.
And always happy to plug ASAS-SN <https://www.astronomy.ohio-state.edu/asassn/index.shtml> (The deleted @SuperASASSN was one of the best astro follows on twitter some years ago), who have found an awful lot of detached eclipsing binaries: <https://academic.oup.com/mnras/article/517/2/2190/6695108?lo...>.