Late to the party I know usually just a lurker..
I have been using Perch since V1, lone developer/designer servicing small businesses in a niche market, in some ways probably the persona Perch was initially conceived for.
Over the years I have accumulated a fair few Perch licenses, and I am certainly guilty of reusing them, I never assign licenses to the client, they remain mine, their sites hosted on my server, the client has the rights to them as long as they remain a client or can choose to pay for it if they go elsewhere (which rarely happens).
Recently however I have moved away from using Perch, for a few reasons, some of which may be down to my personal idiosyncrasies, some maybe not, but anyway for any new licensing model possibly relevent.
1. Perceived lack of development. Already dealt with by plenty of others, it is a major concern, given that Perch now seems to be no more than a side project.
2. Updating. I regularly update Statamic, Craft and ExpressionEngine sites by a simple one click process in the control panel, with Perch I need to maintain a git repo and use a deployment service, otherwise maintaining many sites I will be in for hours of work, and even if paid for is costly for me in terms of time required. This in 2020 seems archaic.
3. Control panel look/feel again this may be personal but it feels dated, and honestly has done for a few years.
4. Perch shop: This seems beyond crazy and a major mistake that this is not a paid add on. My guess is the support it requires has been a large factor in creating the current situation, anybody requiring e-commerce really should be prepared to pay for it, and pay well.
OK this all sounds negative, I do like Perch and would love to continue using it for certain clients, but any new licensing model which inevitably will be more expensive (and I appreciate DEV's have to live) has to come with some benefits for the end user in terms of a vibrant development regime, full time or otherwise.