I really do encourage you to experiment, I just want to give you as much information as possible. I'm adding in this sentence after looking at what I've just written below (I'm on summer break and have way too much time on my hands!), and I just want to reiterate the fact that although it may come off as negative, I really really do encourage everyone to experiment and mess around, and hopefully, prove conventional knowledge wrong in some way (but it's always good to know whats already been learned by those before us too!).
The rules of breeding that apply to most common crop plants, such as tomato or corn do not apply to bananas, even within the wild species, especially since when breeding with other crops you can have a rather wide genepool of closely related plants so you generally will get good compatibility, but with bananas, many wild species will cross but not combine well and you will end up with sterile plants, and of the ones that will be easily compatible with each other there is not much available in the trade, so that is why I recommend working with the Rhodochlamys because there is a decent diversity but they are all closely related enough that you will be able to cross them easily and possibly get them to form F2 and beyond seeds. If you're really persistent, with enough trials and backcrossing you may even be able to form a fertile homozygous hybrid (kind of like Musa ornata, which is actually a very compatible hybrid between M. flaviflora and M. dasycarpa).
Regarding them being diploid and thus seed producing, this is generally true but there are some things you should know (there actually are fertile triploids, and the reason triploid bananas don't have seeds is not related directly to their ploidy level. There are seedless diploids, triploids and tetraploids). Often times what happens with interspecific banana hybrids is that you will get sterile plants, since you would be working with wild species they will not be parthenocarpic, so if you end up with sterile plants (either from being female sterile or there being a lack of viable pollen, either or both are possible), you will end up with nothing, basically just empty skins. Pollination in general is always an issue for homegrown seeded bananas, you will need multiple plants flowering at the same time (the plants are monecious but the female and male flowers are present at different, non-fully overlapping times) and either have some good pollinators around (possible, but more often then not seeded bananas go unpollinated out of their native habitats), or more likely, you will need to hand pollinate the bunches.
Regarding mashing them thru a strainer, seeded bananas tend to be very small and mostly seed with a thin pulp covering the seeds, sometimes in certain species there is some substance to the pulp but even then, they are nothing like other cultivated seeded fruits. In banana breeding, they actually do use a masher-strainer device, but in that case they are doing large parthenocarpic bananas with a few seeds (at most) scattered throughout, some wild bananas can have upwards of 300 seeds per fruit and virtually no pulp. In fact, one species in particular, Musa balbisiana, has so little pulp and is so dry that a peeled fruit can completely dry out and hold form around the seeds before it even has the chance to rot. Part of my studies include collecting banana seeds from wild bananas and I have found that the masher-strainer concept does not work. The best ways I have found to separate the seeds from the pulp is to take a bite, add lots of saliva and mash it around in my mouth for awhile, bit by bit, spitting out the seeds one by one and cleaning them even further after that. The other method I use involves a screen, but instead of trying to mash the pulp the thru leaving the seeds behind (which does not work), I use it to mash the fruit against with my hands while running water over it to separate everything out, then I pick the seeds off the screen out by hand. I use different methods for different species (depending on how moist and tasty they are), but in both cases I cannot recover any substantial amount of pulp.
The pulp content of a wild seeded banana is not comparable to if a normally seedless banana had seeds. This is because in seedless parthenocarpic cultivars, an autonomous auxin release in the fruit causes a layer of tissue just under the skin the expand and fill the ovary, in seeded bananas that auxin signal is not there, and if fertilized the ovary will instead be filled mostly with seeds, if there is no fertilization then the ovary will remain empty.
One may be thinking then, that the ideal compromise would be to have a plant that is both fertile and parthenocarpic so you could have seeds and a decent amount of pulp at the same time. These types of bananas do in fact exist, and they are categorized as primitive cultivars (as opposed to advanced cultivars which are parthenocarpic but completely sterile). The problems with using primitive cultivars in your case are that 1. they are hard to find on the consumer level (like if you want to buy some from a nursery), and there may not be any available that are both male and female fertile which you would need for the type of breeding you want to do and 2. when parthenocarpic cultivars are used in breeding, even if they are diploid, there is a number of poorly understood meiotic errors that can occur and you can actually end up with progeny ranging from diploids to septaploids (on top of heterozygosity) and thus you are back with having to deal with wildly diverse seedling progeny.
The diagrams below (though not comprehensive) will help to illustrate some of the issues I've address above. From: Simmonds, N.W. The Evolution of the Bananas. London: Longmans, 1962.
The above caption which is partially covered by the watermark reads "Cross-sections of fruit pulp shown stippled and seeds solid black; arrows show the consequences of failure of pollination"