We are certainly not opposed to this, although some of the more fundamental structural changes in the codebase might temporarily hurt the stability - this was one of the friction points with the mainline. While stability is important it was never clear where sound judgment ends and "don't touch this 14+-year-old code transpiled from Pascal" begins. We are almost a year in and among several former contributors we still sometimes have troubles figuring out the logic behind certain decisions. Now imagine doing all the digging and expecting to fight an uphill battle to prove the necessity of the changes. Maybe if we do it in the nimskull then mainline decides to adopt some of the improvements after seeing their real worth.