I'm not sure you've thought about the costs and noise associated with cheap space heaters.

When I compare the amount of heat generated by an antminer S1, drawing under 500W, and the about provided by a quiet alternative, an oil radiator, at 1500W....A great many of the space heaters you'll see used in economically disadvantaged homes are ancient radiant heaters that consume 1500W constantly, and are loud as hell. The miner puts out many many more BTUs per hr per watt. I do not have the specific data, but I can assure you it takes less wattage to heat our homes.  

If you consider the amount of $ spent to heat the home, you'll see that the mining gear need not be "profitable" to achieve a net gain for all the parties involved. It only needs to come to a near break-even point to be successful in my own eyes. If you're trying to achieve ROI in a space heater, I believe that you're not trying to solve the correct problem. Remember, you're a miner trying to make a profit mining.

the problem to solve, truly is to heat a home in a financially more efficient way. If there's a reduction of the home users' power bill because the miner recoups some of the cost of generating heat by generating coin, is that not the actual ROI?

Finally, you must also consider, this is but one method of repurposing older gear. There are many more that I'm currently aware of. We are not limited to single ideas on how to improve the network while improving people's lives. Nor are we limited by the short-sightedness of unrealistic expectations of instantaneous ROI that is so prevalent in the cryptomarketplace. Instant ROI simply isn't real in the long term, and is currently a function of novelty and adoption rate. The idea that you can buy miners today, and profit tomorrow is fading quickly. Please note the endless series of scams, scamcoins, and so on that point to frustration with the length of time that it takes to achieve ROI mining. In most businesses it takes 2-3yrs to start realizing a profit. Miners expect a profit in 60days. This is not sustainable.

As for "51% abuses" - in reality one need not have 51% of a network to attack it. In practice, the primary "attack" performed against most blockchains is one that even bitcoin is still susceptible to, and that is "selfish mining" - there are various ideas on how to solve that problem, and the Mazacoin Development Team is actively working on the issue. Selfish mining only needs 25% of the network to be effective, and we see this in bitcoin everyday.

It's my opinion that the majority of algorithms in use today have the same issues as sha256. Each has grown, CPU-only coins are GPU, with FPGAs on the way, and soon to follow will be asics for the most profitable algorithms. Certainly mazacoin will need to grow and adapt, but there is not a reason to change the sha256 algorithm to deter attackers, or selfish miners - these problems are present in each algorithm, and each PoW scheme devised at this time. Multi-algorithms simply increase the challenge to would be attackers and selfish miners, but don't solve the fundamental issues.