This Mastodon page for reducing congressional districts in size made me wonder if your long advocacy for the power of caucuses in Parliament would favour smaller ridings in Canada. Could a house of MPs each representing fewer people, and more closely influenced by them, in turn expect and demand greater influence on party leaders? And would the greater number of MPs, conversely, more readily make their wishes felt?
Short answer: sorry, but I doubt it.
The story Mastodon refers to is this post by Thirty-Thousand.org (an American lobby group), quoting a New York Times editorial in favour of a larger House of Representatives, which has been fixed in numbers since the 1920s.
There's a political science observation that sizes of legislature around the world tend to group around the cube root of the populations they represent. Sometimes this transmutes into a judgment: the number of legislators should match the cube root of the population they represent. By that rule the US House of Representatives (435 seats) should have 695 seats. By the same rule, the Canadian House of Parliament (338 seats) should have ... 335. That is, Canada obeys the cube root "law" very precisely. The Thirty Thousand group thinks the US House should far exceed the cube root number.
So there is that.
On the other hand, there is a longstanding maxim among Canadian political scientists that does seek to explain backbencher docility by the "smallness" of the Canadian House and its caucuses. It proposes that because so many caucus members are either in cabinet, or are parliamentary secretaries, or are aspiring to be one or the other, the remaining backbenchers toe the line out of personal ambition, in hope of future advancement. Political science texts back at least to C.E.S Franks's Parliament of Canada in the 1980s find a causal relationship in the observation that in Britain, where caucuses are larger, deference to leaders is less common. Note that the British House has many more MPs than the cube root law would suggest. I.e., it has the reverse of the American situation, which somehow also produces rather independent members quite often.)
With regard to the question about backbench power, there is really not much evidence to suggest the correlation "larger caucus = more independence" has an empirical basis. And British cabinet ministers, so loyal in theory, are normally as deeply implicated as backbenchers in British caucuses' reviews of the party leadership.
The rule that British caucuses mostly hire and fire their own leaders, while Canadian caucus mostly do not, seems to me much more causative of backbencher independence than some theoretical ratio of MPs to constituents. That system presumes accountability. Our Canadian political culture constantly reminds our elected representatives that they have no authority to disagree with a leader who has been chosen by a self-appointed mob from outside the caucus. That still seems to me much more consequential. Surely it's an issue of political culture in Canada, not a numbers game.
So it's mostly my feeling that as long as backbench MPs are basically useless (from an accountability perspective), there's probably little point in further proliferation of them.
I note that the cube root rule would give Ontario 246 MPPs instead of 124 -- does that seem like a good idea?
Update, December 9: There's now a substantial addition to this conversation here.
