Tankless by Rinnai or noritz only. (I'm a Rinnai guy)
Repairs = Easy if you take a little time and get trained up a bit. No it's not a DIY job.
Boiler/Indirect? Ok, great if I'm in the NE, but what about when it's summer and I have to run a hot BOILER all day to keep my Indirect hot ALL DAY? how about the HEAT GAIN that now my A/C has to overcome from the boiler and the tank to keep the house cool? Nobody ever thinks about that one.
Yes, The DHW is almost free in the winter, but I think you get to pay it back in the summer from several different directions.
Sure they do- all the time! (The efficiency nerds behind CA Title 24 are all over this one, and gawwwll-leee, the subsidies for indirects in oh-so-cold CA are the same as for a tankless.)
But the keeping the boiler hot all summer is why internal coils in boilers SUCK, and indirects are GREAT. You
don't have to keep the boiler hot all summer with an indirect- you fire it when the indirect calls for heat, then let it cool off. With a high-mass boiler it pays to have a higher-mass indirect and set it up for a very large hysteresis to keep the fractional heat abandoned in the boiler small, and guarantee long enough burns that you're not murdering the boiler with cold cycles. But overall it's not much worse than a standalone tank heater in summer, even with a monster-sized cast iron beast. With smaller cast iron boilers it'll beat a standalone tank in summer. Even oil-fired beasts are cold-start tolerant these days.
But if it's a low-mass boiler like a mod-con the boiler itself is usually smaller burner-wise than a whole-house tankless, better INSULATED than a tankless, and spends far less time idling at temp since it's not being brought up to temp every time somebody washes their hands or rinses out a washcloth. That increased efficiency translates into a LOWER load on your AC than with a tankless (provided the standby loss on your indirect is well controlled. Some is, some ain't... but all can be easily brought up to snuff with additional insulation if need be.)
The air conditioning load of a 1/2 degree/hr 30 gallon indirect tank by itself is 125BTU/hr or 3000BTUs/day (and then only if it's maintained at 180F! It's half that if kept at 130F). Call it 4K/day for a 40 gallon indirect.
Care to take a stab at the daily air conditioning load of a single square foot of south facing glazing is in your neighborhood? How about a square foot of skylight? If your neighbor's house is painted white, THAT could represent a higher AC load on YOUR house than a mod-con + indirect!
It wouldn't surprise me if a tankless + indirect combi beats a standalone tankless on minimizing AC load in summer, provided the indirect & plumbing is well insulated.
If you want to really get picky, measure the AC load difference between direct-vented sealed combustion tankless is vs. a tankless that uses room air for combustion. (I'd bet that's the same order of magnitude air-conditioning load as the standby loss of a good indirect.) All mod-cons are sealed combustion/direct vented, which would not be true for standard-efficiency tankless units (if true for condensing versions.)
Don't get me wrong- I have nothing against tankless hot water heaters, but the whole air conditioning load thing with indirects is completely out to lunch. Methinks you're being misled by the issues surrounding tankless coils in boilers, which are a whole other animal, and run at roughly half the summertime efficiency of the same high-mass boiler with an indirect.
For a comparison of indirects vs. internal coils, see:
http://www.nora-oilheat.org/site20/uploads/FullReportBrookhavenEfficiencyTest.pdf
This study models mostly oil-fired WAY oversized cast-iron boilers, which are way less efficient in summer than the oversized gas-fired mod-con they tested (system #11). Smaller units will uniformly test better in water heating mode than bigger units, all other factors being the same. The smallest (90KBTU-in) cast iron oil-fired boiler was tested in both indirect mode and tankless coil mode (systems 12a & 12b) had more than twice the efficiency in indirect-mode than when using it's internal coil for the hot water (24% vs. 51%) , and comparable to a standalone gas-fired tank. (51% vs 57%). The oversized mod-con + indirect made 59%.
In the real world an 0.82-0.84EF Noritz only runs ~70-75% water-heating efficiency due to the short cycling. And while tankless standby losses are low, they're not lower than a mod-con's. Both are zero during long periods between firings, but the time between firings is longer for the mod-con, and the mod-con is better insulated. If you add the AC load of a non-condensing Noritz's combustion-air infiltration aspect to it's inter-draw losses on hand-washing cycles, etc, are you REALLY sure it's less of an air conditioning load in summer than a right-sized 83% AFUE cast iron boiler + indirect? Less than a sealed-combustion & right sized mod-con + indirect?
Maybe so, but I'd need a measurement to verify that.
And they're all a lower AC load than a 10 square feet (one window) of south facing glazing (even the low-E stuff) in most of the US.