My recommomendation for the replacment is...
First: Do a real heat-loss calculation on the place, and verify it against degree-day data and fuel bills. If the rated output of your old system is oversized by 2-3x (all too typical) based on running 100% duty-cycle at design day temps, (which you should figure out by your fuel use per degree-day historical data) adjust your old boiler's rated efficiency down 10-15% and recalculate. Most software or Manual-J methods overestimate by 25-30%, so don't DARE add any additional margin yourself. If your heat-loss calc says one number, and your fuel use calc says a much lower number, go with the lower number as the output you need in the replacement boiler.
If you do that there's some chance it'll meet it's AFUE numbers, or at lest come close. Don't do it, and it almost certainly won't. For a good summary view of what being 2-3x oversized does to your annual efficiency & partial-load efficiency, check out the graphs & tables in the appendices at the end of this document (based on peer-reviewed data at Brookhaven Nat'l Labs):
http://www.nora-oilheat.org/site20/uploads/FullReportBrookhavenEfficiencyTest.pdf
(Unit #3 is almost surely a System 2000, or a system designed to operate simlarly to a System 2000)
A more condensed summary based on some of the same work lives here:
http://www.bnl.gov/est/files/pdf/ButcherAachenPresentation.pdf
Then...
There are many good mod-cons out there- Triangle Tube & Peerless both perform well, as do many others. Find one that best matches your calculated heat loss number no bigger than 10% over, and you're good.
But you may want to tweak the system in other ways too, re-zone it to match your actual-use lifestyle, etc.
...or...
This goes well beyond the original question but I'll stick it in anyway as food for thought on the System 2000, and how you can make your system more efficient using similar high storage mass strategies.
Low/medium mass boilers coupled with (highly insulated) high mass
storage is really the ticket, and the key to System 2000 performance. With the high mass storage there's no such thing as a short-cycling. Independent of load there is a significant minimum burn time, which limits the fixed losses incurred on every cycle.
Some designers of low to moderate load low-temp heating systems have done pretty well using much less sophisticated controls, adapting extremely low-mass cold-start-happy boilers aka "tankless hot water heaters", combined with well-insulated buffer tanks. They do A-OK even without the System 2K's heat-purge control & other features. (I'd really like to see how they fare in a Brookhaven type study. I'm guessing if they make their EF numbers in 10 gallon draws, they do pretty well with longer burns & fewer cycles as long as the return-water temps are reasonably low.) With radiation-returns properly plumbed to create at least some stratification in the buffers, using the tank as the hydraulic separator they can gain significant modulation efficiency under mid & high load conditions, while the high-mass keeps efficiency from falling off a cliff under the unavoidable low-load & intermittent burn conditions.
The whole notion in the mod-con world over the past decade has been that you can match the output to the load with continuous circulation, modulated flame & outdoor reset control. That almost works, but only in non-micro-zoned right-sized systems. When the outdoor temps are 20% of design-day load (a real heating day) and the system is getting a call for heat from a zone that's 10% of the whole house, you're servicing a load barely 2% of the full-on design day load (!). Mod-cons only turn down to 25%, so you're serving a load that's still under 10% of the turned-down output- it's GUARANATEED to incur a signficant cycling loss on that call, cutting into it's net efficiency.
The smallest mod-cons only crank down to 8-10kbtu, which is many times the heat loss of a 200 square foot reasonably insulated room on a 40F day (and those same tiny mod-cons are UNDER sized for running an indirect-fired hot water tank.) You can only really cut into the numbers of cycles & cycling losses only by adding mass to the system. Minimum burn times programmed into the mod-con controls may keep it falling into the efficiency abyss, but not with the same assurance or to the same degree as adding mass to the system.
40-120 gallons of buffer in the system with a reasonable amount of hysteresis to the storage temps decimates partial-load condition cycle numbers. Sure, there are standby losses with any tank, but those can be made almost arbitrarily small with more insulation. Beyond buffer standby the primary efficiency factors become the boiler's steady-state thermal efficiency, the fixed losses from flue purges, and the standby loss of the small amount of water in the heat exchanger (which is very tiny indeed in a tankless HW heater). Condensing HW heaters have thermal efficiencies similar to that of mod-cons, but even mid-efficiency tankless HW heaters will beat their low-mid-80s EF numbers in a buffered heating system configuration.
As heating loads for houses get ever lower the instantaneous loads from hot water heating with an indirect drives boiler sizing more than the actual space heating load. Sizing a mod-con with only 4/1 turn down for an indirect isn't necessarily the best or most-efficient option. While the peak load is the indirect, the bulk of the fuel use is still in space heating. Designing the full system for optimum efficiency isn't trivial. Bigger-better buffers (or buffers with internal DHW heat exchangers and heating systems designed to run at DHW temps or lower) are much under-utilized in the hydronic heating trades, particularly on small-medium sized residential.
A lot of the sophistication of mod-con controls are of-necessity wasted in the field- there's only so much the boiler itself can do when the loads are low, and the system designer makes or breaks it. System 2000 type systems designed with inherent high mass storage and low-mass boilers make it harder to screw up, but I'm sure there are still some folks out there up to the task, eh? ;-)
This white-paper summary of what buffers can do for about ANY zoned heating system is well worth reading for any residential hydronic designer, and explains a lot how the System 2000 can maintain much of it's full-fire efficiency and even beat mod-cons under partial load conditions:
http://www.patkelco.com/uploads/files/4bf77cff8bc8411bb00c67e962d0ef7d.pdf
You can use cast iron, mod-cons & modulating tankless HW heaters with decent sized buffers for similar gains in partial-load performance. "Right-sized" per manual-J bang-bang cast iron boiler performance usually gets killed under partial load and they often fail to meet their AFUE numbers in the real world. But even 350lbs of buffer w/10F of hysteresis can work wonders for lengthening cycles. Many cast iron beasts don't reach their steady-state efficiency with burns shorter than 10-12 minutes, and only see burns that long after overnight setbacks (even on design-day), and end up maintaining system temp with multiple burns that are but a fraction of what it takes to reach steady-state. Anything that adds a significant fraction of the time-to-steady-state burn length makes a measurable difference in system efficiency. (Even using a buffer-mass times hysteresis equal to 1/4 of the time-to-steady-state burn counts, big-time.)