Hi - I thought I'd put in a couple of points that might be of relevance and hope that it gives you the confidence to pursue the project more given that you've been around DNN for quite a while, as have I.
I doubt you'll find much public information on the full depth of capabilities of DNN mainly because when taken at that level, it's an IT department not overly likely to invest heavily in sharing their IP and techniques with the world, or even as simple as where would you discuss things like this?
From the point of view of your thoughts about how deep and how much you can do with DNN as a SaaS model, we've built an environment last year, completely using the Community version of DNN, with two separate instances, not 1, which might be worth considering from your point of view.
The primary website is public facing website with the membership base of 26,000 live members out of a membership of 60,000 members, some who have come and gone. The authentication at this point in time is using a third party provider, not DNN based, but I am reviewing another DNN based membership solution to replace, should the opportunity arise.
We have the branches set up as standalone sites, which there are at this stage over 145 individual portals on the same codebase, and we have been able to provide two levels of admin - one single sign on to all, and one branch administrator.
We have a series of templates, some free, some premium, and they are available to the branches who wish to make the investment, and there is an element of consistency in branding throughout to maintain a global look and feel for branding and links and yet give the local branches their own features and it has been well received.
Because I'm very particular about the modules I will use on my server, including who the developer is and their vision to ensure commercial viability in the long term, we've become very savvy working with our 'core selection' and pushing the boundaries on that.
While I'm happy for our team to take credit on the implementation of these solutions, we are blessed with resellers whom we manage and they have the ideas and we build them. Crazy but true - I have one special client who has over 70 clients he manages at one level and we manage to build and keep them all working and running like clockwork. This has allowed us to stay behind the scenes and come up with ways of making DNN a truly unique experience with the best long term management options I have seen on a product.
In reference to your vision on having WebApi - I think that's a good idea to plug into third party services to automate or embed things that are simply not viable, practical or needed to rewrite. Given that we also write some amazing Linux software we've been able to implement custom functionality that bridges the shortcomings of DNN.
Before doing some of these large projects we do due diligence and check out other solutions and software to make sure we're not rewriting the whole solution and I cannot tell you in enough words how comprehensive an 'out of the box' DNN is compared to other solutions. It's important to not be shop blind or naive about other products,whether it's an established solution or up and coming but there certainly is a benefit to knowing a product well to make some good comparisons.
Each one of your comments which make you want to stick to DNN are valid a good reason and on that basis I think it's worth investigating more.
Some people think that you need to go to the professional version for the extra features but we've been able to build an excellent infrastructure without that investment. I'm not saying that would be the right direction for you, but I can assure you that the community base will take you a long way and we've got sites getting close to 1million hits per month without issue, and at times hundreds logged in at once. (rarely but at certain times of the year it is the case).
It's good to see this type of post on what people want to do online and I hope this gives you more confidence.
Nina Meiers