For additional learning consideration, in the real world, it's a pain to expand address space for a subnet. So, especially, when working with small address spaces, it can be beneficial to use subnets sized such that they provide room for adding hosts. Admittedly, getting this "right" is difficult, because you probably don't need to allocate a /24 for 5 hosts, yet if your allocation begins with only no or only a very few unused addresses, it's annoying to have to reassign network addresses just because you need just one additional IP.
When working with IP allocations, it's also often beneficial to have an overall IP address space allocation plan. This to avoid willy-nilly address space allocation "fragmentation" due to using different size subnets back-to-back.
For example, although most would NOT start (as you didn't) with allocating your 1st /27 followed by a /30 followed by another /27 followed by the second /30 followed by your 3rd /27, etc., if you can "see" why you don't want to do that, they you should also see, ideally, why we don't want to "accidentally" get there either.
The way to (somewhat) avoid that, is allocate a larger block of IP address space for your /27s, for your /28s, for your /30s, such that within the larger scope, there's set aside space for similar sized address blocks. Again, getting this right, is like getting sizing a subnet right. For example, starting with only 3 /27s, you probably don't want a block to be able to allocate 512 /27s, but conversely, and for example, to allocate another /27, it would be after your last /30 leaving, of course, an "odd" sized non-binary address block.
Just as an example, given your 3 /27s we might "set aside" 172.16.0.0/25 for them, allocating /27 subnets from the low end. Then we might also set aside 172.16.0.128/25 for your 3 /28s, allocating /28 networks from the high end. Doing this provides us one available /27 in the first /25, and 5 available /28s in the second /25, but if we needed two /27s we could "grow" into the second /25 or if we needed six /28s we could "grow" into the first /25. Together, of course, we've set aside a single /24 for /27s and /28s.
As p2p networks don't usually grow in address space (per network), but can grow in networks, I've often seen a /24 set aside for all p2ps, whether /30s or /31s.
Again, nothing really wrong with your network allocations, it's a fine classroom implementation, I just mention the above to think about.
