JM, a few thoughts to ponder:
How many Rinnai's do you install per year? Does not the Rinnai have better water temperature control and higher flow rates? Doesn't it cost a couple hundred bucks less and have better tech support?
For efficeincy, if you have a car that gets 30mpg and one that gets 35mpg, and they are both in the driveway "OFF" which one is getting better milage?
Now I get the PVC vent thing, however if the venting is up and out the Rinnai is still less expensive with an easy push joint vent kit and a much better hot water source. These things only run for minutes a day, the rest of the time they are off. I think 90+ is over rated for tankless. Seems to me that tgemperature control and flowrate matter more.
80+ is (quite literally) overrated for most tankless in most installations due to the particulars of the EF test: 10.2 gallon minimum draws are nothing like real-world use, and the short-cycle losses for sub-2 gallon draws erodes efficiency dramatically. In real-world draw profiles 0.82EF type tankless units hit in the mid-70s, condensing versions scoring 0.90+ EF numbers run in the high-70s/low 80s. The higher thermal efficiency isn't worth (much) extra for. (Indeed, you'd get better return in fuel savings on drainwater heat recovery on the main shower for the price-delta on the high/mid efficiency units themselves, vent costs not included.)
Steady-state thermal efficiency is a theoretical upper limit, but not a good measure of the real-world performance. Those with tight controls over flue purge cycles losses will do better than those that don't. Those with small header tanks to mitigate cold water sandwich issues that also inhibit fire when the header is above a minimum temp will also suffer fewer cycles, and thus lower purge-losses from multiple low-volume draws, etc. Steady state thermal efficiency or raw combustion-efficiency numbers are only relevant for very large or continuous-draw duty (pool heating, anyone?), not domestic hot water for typical household use. In space heating, commercial laundry/car-wash, health-club showers, or home solar-backup you might beat 90% with a condensing tankless, but not very often (if ever) in straight-ahead DHW apps.
BTW: A a question for any of you who have taken (or will take) the tech courses from various tankless vendors: Is there a modulation level where they typically peak out on raw combustion/thermal efficiency? Do any vendors share any of that data?
Condensing versions likely peak at lower fire in 25-35% of full-fire range (like most mod-con boilers), but I suspect the single heat exchanger types may do better at the high range (like copper-tube boilers), and may be under 80% for raw combustion efficiency at the low end of the modulation range, but 85%+ in the mid/high range. I could as-easily believe the designs have been tweaked to provide peak efficiency at some other burn level, but haven't been able to find test data (which I'm sure is quite proprietary when tested by the manufacturers). Can anybody either confirm/disabuse me of those guesstimates?