I agree it needs to be case-by-case. I'm happy with sK here.

I'd agree that they're less important, but I'm a little uncomfortable with making forms search-only when they're not variants of one of the visible forms (i.e. 異体字 or 旧字体) or irregular (e.g. 変換ミス). Users who come across 礎案 and look it up might be confused if it doesn't appear anywhere in the entry.
That being said, 礎案 is so rare that it probably *should* be considered irregular. A Google search returns only a tiny handful of real results. I do wonder why the shinmeikai editors felt it was worth mentioning.
On reflection, sK is probably fine here. But I think we should handle these 〜とも書く forms on a case-by-case basis.

@@ -11 +11 @@
-<ke_inf>&rK;</ke_inf>
+<ke_inf>&sK;</ke_inf>

I don't have any objections either way, but I marked it [sK] because I thought that these sorts of peculiar kanji forms found exclusively in smaller kokugos (shinmeikai, meikyo, obunsha, etc.) were not as important as forms found in more authoritative sources (koj, daijr/s).

See for example 霑す in entry 1341810

As shinmeikai doesn't mark it as irregular/incorrect, I don't think it should be sK.

@@ -11 +11 @@
-<ke_inf>&sK;</ke_inf>
+<ke_inf>&rK;</ke_inf>