Mark- in-re DIY drainwater heat exchangers using soft-copper:
Key to getting the heat transfer efficiency is contact area between the potable and the drain. The very original prototype designs in the early 1980s did round 1/2" soft-copper wrap and dipped them in solder but those were expensive to make, and not as efficient as 2nd & 3rd generation units that used flattened potable wraps tempered for some springiness, pre-formed to suitable flatness, then and heat-expande to be able to slip it over the copper drain. It has to be a sufficiently controlled process that a gas-tight contact was made between the flat-formed potable and the copper drain to minimize degradation of the thermal transfer efficiency over time from oxidation crystals pushing the potable away from the drain pipe as slightly insulative oxidation layer grows.
Up for self-engineering all that?
Of the newer designs Renewability does multiple rectangular cross section parallel potable wraps brazed on to manifolds at the connections to the distribution plumbing. All others use either square-ish cross section potable wraps or D-section wrap with the flat side up against the copper drain. Both work pretty well at single shower flow rates, and 4' or shorter units. With higher flow rates & longer units Renewability units have a slight edge from having less of a pressure drop and less thermal mass and higher turbulence in the potable wrap from the multi-path ribbon-pipe.
These things are dumber than a box o' rocks (even the Canuckistanis can manage 'em! ;-) ), and in no way comparable to the short-cycling smoke with mirrors represented by overwrought tankless HW heater efficiency claims. They're zero maintenance, and really DO work as-advertised in the real world if you mount them reasonably plumb. (At least within the measuring accuracy I get on the limited number of field-unit's I've measured with IR thermometers. I'm on year 5 with the the example in my basement.)
The inability to measure the return from a tepid toilet flush doesn't mean they considered it a joke. They are subsidizing it to the amount of net lowered cost achieved by the other ratepayers/taxpayers all over the country to where it's revenue neutral, based on the econometric models of the lifecycle reduction in demand these things achieve, much as other efficiency subsidies are done. If they had been able to measure tepid-water heat recovery that would have gone into the subsidy analysis, but since they couldn't, they stuck to the performance that they can actually measure. And since 2.5 gpm is actually a bit higher than the average real-world shower flows, the real world efficiency is slightly ahead of the NRCAN test numbers.
I don't put any stock in vendor hype about payback & performance, but I DO have some faith in the certified test numbers listed on the NRCAN site. How you use the thing and your actual hot water heating efficiency & fuel cost determines the true payback, but the NRCAN numbers are good enough to design systems for minimum burner size & storage volumes required meet showering loads.