Consider that for new users MeldaProduction is a daunting and complicated world to dip in to (but with lots of reward if you invest the time and money). You have years and years of growth, and new products, and redundancy, and complicated layers upon layers – all good once you understand them, but why not simplify where possible without losing all the benefits?
This process of growth… then trimming the excess… repeat… is important, in a lot of things in life.
Will be painful in the short term, but could have big advantages in the future (for both new and old customers).
Your only hope in achieving this is to make bold moves, ripping the old plaster, but supporting the transition through a parallel phased approach, with clear deadlines:
- Introduce a completely new product line – go big – maybe all new naming schemes (if you want), cut out redundancy, simplify the automation, etc. – make MeldaProduction 10x better than today. But treat it as completely new and improved – free to innovate and simplify the way you want.
- All new Melda products are only developed for the new line.
- Maintain the current product line for a fixed period of time – people can continue using these. Provide critical and important bugfixes, but encourage use of the new product line in new projects. Old projects can continue as normal, but don't need to migrate any existing plugin usage (no real benefit in the short term).
… very similar to what you've (Vojtech) already suggested in your initial post, but more clear on expectations, e.g. migrating existing usage of plugins is not supported (just too much work and hassle for potentially little gain?)
Draw a line in the sand. Use it as an opportunity to take things in a direction you want, for the future.
Or, you know, continue as-is
There's no true single answer, just choices 
)
Damn... The problem is that for sake of just changing the names and keeping the automation parameters intact, one would have not 100 Melda plugins there, but 200... That doesn't seem like a good idea... Perhaps the best way would be indeed to create some backward compatibility option for those, who are dependent on the automation and keep the plugins the way they are... Damn... and I so wanted a clean slate 
). Keep thinking about it for another year and when the time comes you can probably approach the initiative with a good deal more confidence and have really good answers to concerns from your user base.