Well concerning my recent experience at Armorcon, and some of the grumblings from other entrants at least in the AMPS system, maybe anyone who wins a best-of is elevated to Masters Level?
In the AMPS system if you can start at any level, but it is recommended that you start at the beginner level. If you earn a Gold at that level you advance to intermediate (at a national show, regional advancement is optional and maybe it shouldn't be?). Earn a gold in intermediate advance to Advanced. And that is it unless you win a BOS at the nationals, then you move to Masters and only compete from then on in the Master's class. They are still eligible to win best-ofs, but not BOS at the nationals.
I see a couple of problems here.
1. Advancement is only mandatory at the Nats. This can leave a good modeler in a group they shouldn't be in. Now in AMPS you really don't compete with others in your class, you compete against a theoretical perfect model. How you do, has no bearing on how others do, so everyone in the class can win Gold, or no one can. So maybe this is a mute point. In the AMPS system the only time you compete against others is for BEST OFs. Here the skill level of the modeler doesn't matter, only the overall score. Perfect is 30.5, if you don't have that, you are out of the running.
2. Generally what happens is people who win BOS at the Nats - never compete again. They may attend the shows, but they rarely enter. While this only "knocks out" one a year, we are an old group, and really do we want to do this? Maybe a modeler who wins a Best of or BOS at any show, regional or national is advanced to Master. This would create a larger pool, and maybe we would see less attrition. As masters only compete in the masters class - ie no DIO no Vignettes, not open top halftracks etc, what they bring goes into the master's class. This allows others to win Best of and prevents sweeps. Masters could win Best of Masters and still would be eligible for BOS.
This is what comes to mind right now, I am sure I will think up others. As for IMPS - I'm sorry, I see no way to fix that cluster. The last IPMS show I attended these fools were measuring the height from wing to table - there is no hope for them.