Boiler sizing based on total radiation is right for steam systems, but for hydronics, not so much. It's too much boiler, and too high a min-modulation. To get the most out of a mod con, look at the MINIMUM modulation number more than the max- the lower the min, the more time it'll spend in a modulating mode=fewer, longer cycles, more time spent in condensing mode, less wear & tear on the system. The system design also has to minimize the return-water temp to the boiler to wring out that 90% + efficiency. This is (most likely) a dual water temp system, with lower water temp for the radiant than for the fin-tube(?).
With only 100' of baseboard (and another 100-150' of distribution plumbing) with a low mass boiler with a 50k min modulation it'll short cycle on the high-temp zone. (Even a smaller mod con could have issues). I'm assuming the radiant is in a slab, and already has sufficient mass to preclude short-cycling on that zone? You can add mass to the high-temp zones with a buffer tank, but what makes more sense is to buffer the high temp zone with a "reverse indirect" hot water heater (Everhot EA series, Ergomax, Turbomax) maintained at whatever temp that zone requires, with a tempering valve on the DHW heat exchanger's output roughly this:
This would make the return water to the boiler a bit high for good condensing, but even an oversized boiler would now have a minimum burn cycle. If you ran the tank as one zone of a 2 zone system (from the boiler's point of view), running all high temp zones off the tank, and ran the radiant as it's own zone (not running off the tank) you can maximize the condensing benefit.
But go with a smaller mod con with a sub-20kbtu/h min-mod (Burnham Alpine 80, Triangle Tube Prestige Solo 60, Peerless Pinnacle T50/T80, etc.) and you'll do a lot better efficiency-wise. You basically want something that'll run 100% duty cycle whenever it's below 30F, but with enough of a high end to deliver the design-day heat load, with a bit left over for the hot water. With a design-day heat load of ~45K, a 60-80K max modcon should do it (a 50K will if you have a big enough indirect.)
As a side note: If there's wall space for it, with ~50-75% more baseboard you can usually get the heat out of it at low temps too. If you can get your design temps down to ~125-130F or so you can run the whole shebang off a reverse-indirect maintained at 130F and still get sub 120F return water for 90%+ efficiency.
If you stick with oil, the Energy Kinetics System 2000 uses a medium mass boiler and smart controls to purge heat from the boiler into an indirect at the end of each burn cycle to minimize cycling losses, and it would DEFINITELY beat an oversized mod-con on efficiency:
This is similar in effect to the reverse-indirect buffering with a low mass boiler, but since the thermal mass of the boiler itself is higher, they need the purge cycles to keep it from falling off an efficiency cliff with heat abandoned in the boiler. With a properly sized low mass mod-con and a buffer it'll modulate down under part load rather than cycle, and when the load is lower than the min-mod, the amount of heat abandoned in the boiler each cycle is lower- a tiny fraction of that delivered to the buffer tank, and net efficiency stays high.
There are even some decently efficient designs using this approach using tankless HW heaters instead of mod-cons. You're limited to about 83-84% efficiency with standard tankless HW heater, but unlike bang-bang cast iron boilers, it'll DELIVER that efficiency in the real world even in during the shoulder-seasons:
Even if you went with a lower mass oil-fired boiler, you'll maximize the efficiency of the boiler if you buffer the fin-tube, and using a reverse-indirect kills 2 birds with one stone. There's a lot to be said for integrating the heat/hot water even WITHOUT the buffering effect (running a standard indirect as a "priority zone") but the efficiency gained with the buffering is substantial, especially with heat loads much lower than the boiler's output. See:
(Unit number 3 is a system 2000, system #11 is an oversized mod-con. Skim the whole thing, but note the degradation of oversizing in Table 4, and peek at the part load regression curves in the appendices for a clearer picture.)
For a comparison caluculator based on the test data from the above summary, see: