Too many `nixpkgs` nodes in `flake.lock` file.

[jaredponn]

Is there a reason why https://github.com/Plutonomicon/cardano-transaction-lib/blob/e23ff270b3013db769c0dba1e7cfb284f4517a82/flake.nix#L76 is not following CTL's haskell-nix? I.e., would it break everything if we had

diff --git a/flake.nix b/flake.nix
index a38d4555d..43e39dcdb 100644
--- a/flake.nix
+++ b/flake.nix
@@ -73,7 +73,10 @@
     };

     blockfrost.url = "github:blockfrost/blockfrost-backend-ryo/v1.7.0";
-    db-sync.url = "github:input-output-hk/cardano-db-sync/13.1.0.0";
+    db-sync = {
+        inputs.haskellNix.follows = "haskell-nix";
+        url = "github:input-output-hk/cardano-db-sync/13.1.1.3";
+    };

     # Plutip server related inputs
     plutip = {

instead?

Currently, CTL has about 120 nodes for nixpkgs in it's flake.lock file,

$ cat flake.lock | jq '.nodes | keys' | awk -e 'BEGIN { i = 0 } /nixpkgs/ { i++ } END {print i }'
123

but this suggestion reduces it to

$ nix flake lock 
<blah blah blah>
$ cat flake.lock | jq '.nodes | keys' | awk -e 'BEGIN { i = 0 } /nixpkgs/ { i++ } END {print i }'
36

For context, in lambda-buffers doing anything with nix (nix develop, nix build, etc.) takes >10s because the flake.lock is so large, and I'm pretty sure that CTL's db-sync is the last major contributing factor to this.