usually a g3 london is kinda good-ish for axis, while a g4 london is almost a gg.
the biggest factor why sealion is a bad move for axis is that after a proper usa response (bmbrs), japs cant dow till london s captured, which most of the time (-if not always) means no pac victory for axis. so japs play gets really restricted. might be the only occasion where going heavy into siberia with japs is a more optimal move than going for india.
the general outcome also depends on a number of minor factors:
-can ita grab egy before (or at the same round) usa gets a stack in sz91? with london gone it s almost impossible for uk to send reinforcements to egy + relatively later involvement of ussr means it will take a longer time for ussr to get involved in africa.
-how long can ussr stack romania? romania is the usual optimal base for ussr stack, trading slovakia/yugoslavia plus dispatching some guys to bulg for greece and albania means, after a successful sealion, ger will find itself outproduced by ussr.
-how lucky was ger in london? having a lucky attacks with many land units left means ger can remove ussr from eastern europe. many air left further means ger can probably delay usa fleet london for 1 or 2 more rounds.
-is iraq italian or russian soil? a 9 ipc territory plus enables pressure on cau or egy. no further need to emphasize its importance imho.
with all those things said, I really think sealion is a very “specialized” option for axis, which must be only used with in depth analysis of the board, even when it s possible for ger to take london in g3.
so in return, uk doesnt have to solely turn london into a fortress to deny a selion. depending on overall situation, with careful management, making ger grab london can be a game-winner move for allies on the long run.