Is it possible to build the 100/1000Base-T signal route from the RJ45 to the magnetics, and from magnetics to PHY via throughholed passives. Like resistors and capacitors. If care is taken to avoid stubs, have straight lines, matched lengths etc..
I was thinking on a smt approach. But I had some IRL suggestions it might just be overkill. But as throughhole seems to be gone since the 10Base-T days I assumed (bad thing in science :). That throughholed passives would screwup the signals. 100Base-T operation will certainly be used in this case.
1000Base-T is more like "maybe".