I find that using MatchSrf in V7 is way slower than in V5 - the way that you select edges in V7 really makes it so much more click intensive. For instance - In V5, when you are at the “Select segment to match” stage on the CLI, it gives you a sort of “soft option” to chain edge - that is, you can simply select one edge and hit enter, OR you can select multiple edges. There’s no need to select anything in the CLI, and the user can go either direction. In V7, it forces you to go up to the CLI (or enter “C”) if you want to chain edges. If you forget to do so, the pop-up comes up and if you’ve messed up you have to cancel out and go back and start the command again. This sounds like a small thing, but wow does it add time to using MatchSrf when you use it a lot. Is there way to get the V7 UI/UX the same as V5?
People were complaining about that change years ago, but seems like the developers have their own opinion on this particular case and won’t listen to the Rhino users. If I remember correctly, that was the case for “Blend surface” as well. The current implementation forces you to preliminary specify whether a single or multiple edges will be selected, which brings the problems you mentioned above. Not to mention that there is no option to add extra target edges or curves during the preview of both, “Match surface” and “Blend surface”. Making edge fillets lets you add more edges, but those two tools don’t show consistency in that regard.
…and then @theoutside gives me grief for still doing most of my work on V5 
Another huge issue with both, “Match surface” and “Blend surface” is that Rhino will not remember the last used target edges (or curves), hence upon failed or repeated command to apply better settings the user is forced to pick those all over again. Remembering the last used input for the current session could reduce many mouse clicks and time. 
I also don’t get it why “Revolve” can’t remember the last used axis, while “Rotate” and “Rotate 3d” can? Or, why “Fillet edge” can’t remember the last used edges in case that the command fails or needs to be redone with better settings? Or, why “Orient” and “Orient 3pt” can’t remember the 2 or 3 reference points.
I’m aware that “Named selection” exists, but it’s a cumbersome solution, because it forces you to preliminary select all input edges and/or curves, then save them manually, then bring the “Named selection” panel, then select the saved named selection, then close the panel, then run the “Fillet edge” or “Match surface” or “Blend surface”. A more intelligent solution is to just make Rhino automatically remember the input edges and/or curves and let the user press “UseLast” in the Command line to make the magic.
I know I sound like a broken record here - but the VSR matching command just assumes that you want to match a surface edge to…the one right next to it. So you just select the edge you want to change, and boom, it auto selects the edge to match to and pops up the dialog box. If it happens to grab the wrong one (rare, but happens), you can over ride the selection in the command while it’s still up, instead of having to go back and re-start the command. It does this same thing on “Align” which is their separate command for matching multiple edges of the same surface simultaneously. You literally just select the surface, and then the dialog box pops up, you can adjust the G0/G1/G2 for each edge if you so desire.
This is how it should be in Rhino 7 and Rhino 8 WIP as well, but unfortunately everything is too slow and requires plenty of mouse clicks for executing simple tasks. Maybe we will have to wait until Rhino 9 or Rhino 10? I suppose that the automatic selection of the most appropriate target objects is made by telling Rhino to consider the closes surface edges within a given distance.
That problem with slow and tedious selection of edges also exists in the “Edge continuity” tool that forces you to pick 8 edges to be able to analyze the 4 sides of a single surface! It also can’t analyze one surface edge with two or more nearby edges.
To get the V5 behavior, you can use the macro _MatchSrf pause C
that’s right, for this you can use
_BlendSrf C AutoChain=No
There is a _PreviousEdgeSelection option, when does this not work for you?
There is no such option when I repeat ! _BlendSrf or ! _MatchSrf. Only ! _FilletEdge has it.
I know, but my comment was to your FilletEdge remark, so I wondered if something did not work in that command in some cases.
It works, but my main concern is the lack of such option for both, “Match surface” and “Blend surface”.
I wonder when will we have the chance to choose G0/G1/G2 for each edge when using multiple match?
This would save a lot of time…
If _NetworkSrf has it (a command I never use, because of its fixed degree 3 result), why can’t match have it?