Google N-gram Corpus Counts
| 叱る　 | 286,301 | 45.2% |
| 呵る　 |     285 |  0.0% | <- in daijs, koj, nikk, mk
| しかる | 346,668 | 54.7% |

@@ -12,0 +13 @@
+<ke_inf>&rK;</ke_inf>

𠮟る is still not that common (10k raw Googits vs 4.7M), and can't get into the legacy EDICT versions because it's not in JIS 208/212 (it's in 213).
I'd prefer that it was later in the entry.

From the official 常陽漢字 standards document:
「𠮟」と「叱」は本来別字とされるが，その使用実態から見て，
異体の関係にある同字と認めることができる。

In practice they are the same character.
However, 叱る is the safer choice, it is supported by more text encodings.
(Feeding 𠮟る into the n-gram server I get an "Internal Server Error".)

@@ -5 +5,2 @@
-<keb>𠮟る</keb>
+<keb>叱る</keb>
+<ke_pri>ichi1</ke_pri>
@@ -8,2 +9 @@
-<keb>叱る</keb>
-<ke_pri>ichi1</ke_pri>
+<keb>𠮟る</keb>

Joyo-2010 uses 𠮟 in place of 叱

@@ -3,0 +4,3 @@
+<k_ele>
+<keb>𠮟る</keb>
+</k_ele>