Rankfootball is a fine ranking system. Of course it is much better than the FIFA rankings. However, the "mathematical foundations" of the FIFA rankings are so absurd that it was only RSoccer and Mondfoot, to produce a worse ranking.
Rankfootball may be compared with the rankings, which are calculated on a scientific basis. The operators of AQB (University of Waikato, New Zealand), Chance de Gol (University of Sao Paolo), Soccer Power Index (Nate Silver), or Voros McCracken are all no fools ... eloratings adapted a system of a professor of theoretical physics …
In eloratings is still evident that teams from Asia or Oceania are classified clearly too strong (it can be substantiated statistically). This is not due Professor Elo - but by the arbitrary setting of eloratings start ranking list. This structural problem is not resolve by itself. But - It should not be ignored that eloratings is still a giant among ranking systems.
Other rankings have other structural problems. Rankfootball for example has the problem that the differences in skill level (look to the rating scores) are estimated considerably too low. Assuming small differences in skill level, the result of a single match has a stronger effect on the outcome of the assessment of a team. This is the reason that it can come quickly to major changes in the rankings.
Last example that I noticed was Venezuela. They advanced after the first round in the Copa América up to rank 7 - before they were eliminated by Paraguay and then lost 1-4 against Peru, El Salvador 1-2 and Honduras 0-2. In 2001, Brazil was ranked on 20! Half a year later, Brazil won the World Cup (winning all seven matches without any extra-time).
These are all extreme outliers, and certainly not representative for rankfootball, but these outliers are consistent. An assumed low skill level gap between the teams other hand, leads to the fact that performances in friendly matches are represented fairly well. (There the performance gaps are not as serious, because the favorites of these matches often use them to test new players and tactical systems).
In addition, rankfootball undervalued African teams clearly, which may be due to the overvaluation of results of friendies.
Rankfootball is a nice ranking system! The public likes an up and down in the rankings! Maybe the big moves in the rankings are not realistic ... but normally not absurd.
For the rankings of Roon Ba: it really is overall somewhat more accurate than e.g. eloratings - with one exception, even more accurate (or as accurate) than any other. But the advantage resulting from a greater accuracy with teams who are ranked behind 100 of the rankings. In this "region of the rankings," can only compete Soccer Power Index ... The skill level of teams in the top 100 are shown in eloratings better (despite some glaring misconceptions of Asian teams). This is because the Roon Ba Ranking is calculated on goal difference.
Soccer Power Index, which is also heavily weighted in the goal difference, shows a similar structure: very strong with teams with only a few results. (These teams are accumulated in the lower half of the rankings). In this range the method (calculating with goal difference) works very, very good.
For the rankings of the other 15 sports (and perhaps for women's football) the system of Roon Ba is probably optimal.
Huge strong is also the system of Voros McCracken, which unfortunately is no longer updated and published. Note also the strong Aggregated Football World Ranking!