Having had a chance to think through this...
First, Mark has said that for existing, not upgrading, groups, existing, legacy, price and features will be maintained, so they are unaffected - good for those existing groups, thank you Mark.
For (prospective) new groups, if they are going to be 'Premium' or 'Enterprise', the change boils down to an increase in price, if number of members gets above a certain level. And this increase will (unless member numbers are well managed) be unpredictable - I wonder many groups do manage their member number well, as distinct from just watching members join and leave. But I think any group - or rather those running it (or proposing to) - that has the financial situation to be able to pay under the previous charge structure, will be able to under the new. So not actually a great change in this area.
The big change is for Basic, free, groups, which will, for new groups, be severely restricted in numbers, to 100 members. Many are run by by individuals for, or as, informal organisations, that do not have the financial setup to be able (regardless of willingness) to make ANY payment. I would suggest that a very high proportion of existing such groups were only set up, and continue to exist, because they are free (or should I say zero charge?).
Some future groups will be small, and 'closed', so their numbers will not grrow: so they know that, now and in the future, they will be below 100: they are essentially unaffected, able to be free as before.
The real problem is for new 'open' groups, that may start small, but hope (however realistic this might be), or are willing, to grow, into the hundreds or thousands. And for them, groups.io is no longer making an offer....it will be saying 'we will not cater for you'.
Which will be pity: something that should not be happening. To which I would suggest (as before) that the answer is for Mark to solicit donations, from group owners and mebers, to support free groups (existing and future); as many others (e.g. Wikipedia) are.
A further thought: there is a current thread on GMF 'Determining active members (and removing the inactive)' - to which, response might be summarised as 'not easily' - with memberships being chargeable in future, inactive members will be a cost, and there will be a drive for an easy solution to this problem. (First question - define an incative member)