I'm sure he can make it flexible enough, he just refuses to make it compatible with old software.
That's an ongoing discussion. I personally would be "happy enough" if the core would be powerful enough to at least allow an emulation in software.
The 68040 issue may not be applicable if he's not doing bus accesses in the same way.
There are two problems of the 68040. The first is that it does bursting for cachable regions, which is not acceptable for Zorro-III I/O regions. This is the bus problem you mention.
The second problem is the cache granularity of 16 bytes per entry. The 68040 cannot see modifications of RAM in Zorro space only, and a dirty cache line is always written back completely, regardless of which individual entry is dirty. That is at least the source of the DMA problem and (one) source of the need for CachePreDMA/CachePostDMA().